You are on page 1of 729

The American Psychiatric Publishing

Textbook of
Personality Disorders

This page intentionally left blank

The American Psychiatric Publishing

Textbook of
Personality Disorders
Edited by

John M. Oldham, M.D., M.S.


Andrew E. Skodol, M.D.
Donna S. Bender, Ph.D.
Associate Editors

Glen O. Gabbard, M.D.


Joel Paris, M.D.
M. Tracie Shea, Ph.D.
Thomas A. Widiger, Ph.D.

Washington, DC
London, England

Note: The authors have worked to ensure that all information in this book is accurate at the time of publication
and consistent with general psychiatric and medical standards, and that information concerning drug dosages,
schedules, and routes of administration is accurate at the time of publication and consistent with standards set
by the U.S. Food and Drug Administration and the general medical community. As medical research and practice
continue to advance, however, therapeutic standards may change. Moreover, specific situations may require a
specific therapeutic response not included in this book. For these reasons and because human and mechanical
errors sometimes occur, we recommend that readers follow the advice of physicians directly involved in their
care or the care of a member of their family.
Books published by American Psychiatric Publishing, Inc., represent the views and opinions of the individual
authors and do not necessarily represent the policies and opinions of APPI or the American Psychiatric Association.
Copyright 2005 American Psychiatric Publishing, Inc.
ALL RIGHTS RESERVED
Manufactured in the United States of America on acid-free paper
09 08 07 06 05
5 4 3 2 1
First Edition
Typeset in Adobes Palatino and Optima.
American Psychiatric Publishing, Inc.
1000 Wilson Boulevard
Arlington, VA 22209-3901
www.appi.org
Library of Congress Cataloging-in-Publication Data
The American Psychiatric Publishing textbook of personality disorders / edited by John M. Oldham, Andrew E.
Skodol, Donna S. Bender.1st ed.
p. ; cm.
Includes bibliographical references and index.
ISBN 1-58562-159-5 (hardcover : alk. paper)
1. Personality disorders. 2. Personality disordersTreatment.
[DNLM: 1. Personality Disorderstherapy. 2. Personality Disordersdiagnosis. 3. Personality Disorders
etiology. WM 190 A5125 2005] I. Title: Textbook of personality disorders. II. Oldham, John M. III. Skodol,
Andrew E. IV. Bender, Donna S., 1960 V. American Psychiatric Publishing.
RC554.A247 2005
616.8581dc22
2004023812
British Library Cataloguing in Publication Data
A CIP record is available from the British Library.

To our families, who have supported us:


Karen, Madeleine, and Michael Oldham;
Laura, Dan, and Ali Skodol; and
John and Joseph Rosegrant.
To our colleagues, who have helped us.
To our patients, who have taught us.
And to each other, for the friendship that has enriched our work together.

This page intentionally left blank

Contents
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xvii

Part I

Basic Concepts
1

Personality Disorders: Recent History and Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . 3


John M. Oldham, M.D., M.S.

Theories of Personality and Personality Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17


Amy Heim, Ph.D.
Drew Westen, Ph.D.

Categorical and Dimensional Models of Personality Disorders . . . . . . . . . . . . . . . . . . . . . . . . . 35


Thomas A. Widiger, Ph.D.
Stephanie N. Mullins-Sweatt, M.A.

Part II

Clinical Evaluation
4

Manifestations, Clinical Diagnosis, and Comorbidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57


Andrew E. Skodol, M.D.

Assessment Instruments and Standardized Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89


Wilson McDermut, Ph.D.
Mark Zimmerman, M.D.

Course and Outcome of Personality Disorders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103


Carlos M. Grilo, Ph.D.
Thomas H. McGlashan, M.D.

Part III

Etiology
7

A Current Integrative Perspective on Personality Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . 119


Joel Paris, M.D.

Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Svenn Torgersen, Ph.D.

Genetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
C. Robert Cloninger, M.D.

10

Neurobiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Emil F. Coccaro, M.D.
Larry J. Siever, M.D.

11

Developmental Issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171


Patricia Cohen, Ph.D.
Thomas Crawford, Ph.D.

12

Attachment Theory and Mentalization-Oriented Model


of Borderline Personality Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Peter Fonagy, Ph.D., F.B.A.
Anthony W. Bateman, M.A., F.R.C.Psych.

13

Role of Childhood Experiences in the Development


of Maladaptive and Adaptive Personality Traits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Jeffrey G. Johnson, Ph.D.
Elizabeth Bromley, M.D.
Pamela G. McGeoch, M.A.

14

Sociocultural Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223


Theodore Millon, Ph.D., D.Sc.
Seth D. Grossman, Psych.D.

Part IV

Treatment
15

Levels of Care in Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239


John G. Gunderson, M.D.
Kim L. Gratz, Ph.D.
Edmund C. Neuhaus, Ph.D.
George W. Smith, M.S.W.

16

Psychoanalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Glen O. Gabbard, M.D.

17

Psychodynamic Psychotherapies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275


Frank E. Yeomans, M.D.
John F. Clarkin, Ph.D.
Kenneth N. Levy, Ph.D.

18

Schema Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289


Jeffrey Young, Ph.D.
Janet Klosko, Ph.D.

19

Dialectical Behavior Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307


Barbara Stanley, Ph.D.
Beth S. Brodsky, Ph.D.

20

Interpersonal Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321


John C. Markowitz, M.D.

21

Supportive Psychotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335


Ann H. Appelbaum, M.D.

22

Group Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347


William E. Piper, Ph.D.
John S. Ogrodniczuk, Ph.D.

23

Family Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359


G. Pirooz Sholevar, M.D.

24

Psychoeducation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Perry D. Hoffman, Ph.D.
Alan E. Fruzzetti, Ph.D.

25

Somatic Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387


Paul H. Soloff, M.D.

26

Therapeutic Alliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405


Donna S. Bender, Ph.D.

27

Boundary Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421


Thomas G. Gutheil, M.D.

28

Collaborative Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431


Abigail Schlesinger, M.D.
Kenneth R. Silk, M.D.

Part V

Special Problems and


Populations
29

Assessing and Managing Suicide Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449


Paul S. Links, M.D., F.R.C.P.C.
Nathan Kolla

30

Substance Abuse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463


Roel Verheul, Ph.D.
Louisa M.C. van den Bosch, Ph.D.
Samuel A. Ball, Ph.D.

31

Violence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
Michael H. Stone, M.D.

32

Dissociative States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493


Jos R. Maldonado, M.D.
David Spiegel, M.D.

33

Defensive Functioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523


J. Christopher Perry, M.P.H., M.D.
Michael Bond, M.D.

34

Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
Leslie C. Morey, Ph.D.
Gerianne M. Alexander, Ph.D.
Christina Boggs, M.S.

35

Cross-Cultural Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561


Renato D. Alarcn, M.D., M.P.H.

36

Correctional Populations: Criminal Careers and Recidivism . . . . . . . . . . . . . . . . . . . . . . . . . . 579


Jeremy Coid, M.D.

37

Medical Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607


Peter Tyrer, M.D.

Part VI

New Developments and


Future Directions
38

Brain Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623


Ziad Nahas, M.D.
Chris Molnar, Ph.D.
Mark S. George, M.D.

39

Translational Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641


Martin Bohus, M.D.
Christian Schmahl, M.D.

40

Development of Animal Models in Neuroscience and Molecular Biology . . . . . . . . . . . . . . . . 653


Michael J. Meaney, Ph.D.

41

Biology in the Service of Psychotherapy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669


Amit Etkin, M.Phil.
Christopher J. Pittenger, M.D., Ph.D.
Eric R. Kandel, M.D.

Appendix: DSM-IV-TR Diagnostic Criteria for Personality Disorders . . . . . . . . . . . . . . . . . . . . 683


Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691

This page intentionally left blank

Contributors
Renato D. Alarcn, M.D., M.P.H.
Professor of Psychiatry, Mayo Clinic College of Medicine; Chair, Inpatient Psychiatry and Psychology Division, and Medical Director, Mayo Psychiatry and
Psychology Treatment Center, Rochester, Minnesota

Michael Bond, M.D.


Psychiatrist-in-Chief, Sir Mortimer B. Davis Jewish
General Hospital; Associate Professor, Department of
Psychiatry, McGill University, Montral, Qubec,
Canada

Gerianne M. Alexander, Ph.D.


Assistant Professor of Psychology, Department of
Psychology, Texas A&M University, College Station,
Texas

Beth S. Brodsky, Ph.D.


Assistant Clinical Professor of Medical Psychology,
Department of Psychiatry, Columbia University College of Physicians and Surgeons; Research Scientist,
Department of Neuroscience, New York State Psychiatric Institute, New York, New York

Ann H. Appelbaum, M.D.


Clinical Professor of Psychiatry, Department of Psychiatry, Columbia University College of Physicians
and Surgeons, New York, New York

Elizabeth Bromley, M.D.


Robert Wood Johnson Clinical Scholar, West Los Angeles VA Mental Illness Research, Education, and
Clinical Center (MIRECC) and Department of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles, California

Samuel A. Ball, Ph.D.


Associate Professor of Psychiatry, Department of Psychiatry, Yale University School of Medicine, West Haven, Connecticut

John F. Clarkin, Ph.D.


Professor of Clinical Psychology in Psychiatry,
Department of Psychiatry, Weill Medical College of
Cornell University, New York, New York

Anthony W. Bateman, M.A., F.R.C.Psych.


Visiting Professor, Sub-Department of Clinical Health
Psychology, University College London; Consultant
Psychotherapist, Barnet, Enfield, and Haringey Mental Health Trust, London, England

C. Robert Cloninger, M.D.


Wallace Renard Professor of Psychiatry, Genetics, and
Psychology, Washington University School of Medicine, St. Louis, Missouri

Donna S. Bender, Ph.D.


Assistant Clinical Professor of Medical Psychology in
Psychiatry, Columbia University College of Physicians and Surgeons; Research Scientist, Department of
Personality Studies, New York State Psychiatric Institute, New York, New York

Emil F. Coccaro, M.D.


Ellen C. Manning Professor and Chairman, Department of Psychiatry, University of Chicago, Chicago,
Illinois

Christina Boggs, M.S.


Graduate Student, Department of Psychology, Texas
A&M University, College Station, Texas

Patricia Cohen, Ph.D.


Professor of Psychiatry, Columbia University College
of Physicians and Surgeons, New York, New York

Martin Bohus, M.D.


Chair in Psychosomatic Medicine, University of
Heidelberg; Director, Department of Psychosomatic
Medicine and Psychotherapy, Central Institute of
Mental Health, Mannheim, Germany

Jeremy Coid, M.D.


Professor of Forensic Psychiatry, Forensic Psychiatry
Research Unit, St. Bartholomews Hospital, London,
England

xiii

xiv

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Thomas Crawford, Ph.D.


Assistant Clinical Professor of Medical Psychology,
Columbia University College of Physicians and Surgeons, New York, New York
Amit Etkin, M.Phil.
Center for Neurobiology and Behavior, Kavli Institute
for Brain Sciences, Columbia University, New York,
New York
Peter Fonagy, Ph.D., F.B.A.
Freud Memorial Professor of Psychoanalysis and
Director of the Sub-Department of Clinical Health
Psychology, University College London; Chief Executive of the Anna Freud Centre, London, England; and
Consultant to the Child and Family Program, Menninger Department of Psychiatry, Baylor College of
Medicine, Houston, Texas
Alan E. Fruzzetti, Ph.D.
Associate Professor and Director, Dialectical Behavior
Therapy Program, University of Nevada, Reno, Nevada
Glen O. Gabbard, M.D.
Brown Foundation Chair of Psychoanalysis and Professor, Department of Psychiatry, Baylor College of
Medicine; Training and Supervising Analyst, Houston-Galveston Psychoanalytic Institute; Joint Editorin-Chief, International Journal of Psychoanalysis,
Houston, Texas
Mark S. George, M.D.
Distinguished Professor of Psychiatry, Neurology,
and Radiology, Brain Stimulation Laboratory, Center
for Advanced Imaging Research, Medical University
of South Carolina, Charleston, South Carolina
Kim L. Gratz, Ph.D.
Clinical and Research Fellow, Center for the Treatment
of Borderline Personality Disorder, McLean Hospital,
Harvard Medical School, Boston, Massachusetts
Carlos M. Grilo, Ph.D.
Professor of Psychiatry, Department of Psychiatry,
Yale University School of Medicine, New Haven,
Connecticut
Seth D. Grossman, Psych.D.
Research Associate, Institute for Advanced Studies in
Personology and Psychopathology, Coral Gables,
Florida
John G. Gunderson, M.D.
Professor of Psychiatry, Harvard Medical School;
Director, Psychosocial and Personality Research,
McLean Hospital, Boston, Massachusetts

Thomas G. Gutheil, M.D.


Professor of Psychiatry, Harvard Medical School, and
Co-Director, Program in Psychiatry and the Law, Massachusetts Mental Health Center, Boston, Massachusetts
Amy Heim, Ph.D.
Private practice, Hoover & Associates, Chicago, IL
Perry D. Hoffman, Ph.D.
President, National Education Alliance for Borderline
Personality Disorder (NEA-BPD), Rye, New York;
Research Associate, Department of Psychiatry, White
Plains, New York, and Weill Medical College of Cornell University, New York, New York
Jeffrey G. Johnson, Ph.D.
Associate Professor of Clinical Psychology, Department of Psychiatry, College of Physicians and Surgeons, Columbia University; and Research Scientist
IV, Epidemiology of Mental Disorders Department,
New York State Psychiatric Institute, New York, New
York
Eric R. Kandel, M.D.
Center for Neurobiology and Behavior, Kavli Institute
for Brain Sciences, Howard Hughes Medical Institute,
Columbia University, New York, New York
Janet Klosko, Ph.D.
Codirector, Cognitive Therapy Center of Long Island,
Great Neck, New York; Senior Therapist, Cognitive
Therapy Center of New York, New York, New York;
and Clinical Psychologist, Woodstock Woman's
Health, Woodstock, New York
Nathan Kolla
Undergraduate Research Program, Suicide Studies
Unit, Department of Psychiatry, St. Michaels Hospital, University of Toronto, Toronto, Ontario, Canada
Kenneth N. Levy, Ph.D.
Assistant Professor, Department of Psychology, Pennsylvania State University, University Park, Pennsylvania; Adjunct Assistant Professor of Psychology,
Department of Psychiatry, Joan and Sanford I. Weill
Medical College of Cornell University, New York,
New York
Paul S. Links, M.D., F.R.C.P.C.
Arthur Sommer Rotenberg Chair in Suicide Studies,
Professor of Psychiatry, Department of Psychiatry,
St. Michaels Hospital, University of Toronto,
Toronto, Ontario, Canada

Contributors

Jos R. Maldonado, M.D.


Associate Professor and Chief, Medical and Forensic
Psychiatry Section; Chief, Medical Psychotherapy
Clinic, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine; Medical
Director, Psychiatry Consultation/Liaison Service,
Stanford University Medical Center; Faculty, Center for
Biomedical Ethics and Chair, Ethics Committee, Stanford University Medical Center, Stanford, California
John C. Markowitz, M.D.
Research Psychiatrist 2, New York State Psychiatric
Institute; Clinical Associate Professor of Psychiatry,
Weill Medical College of Cornell University; Adjunct
Clinical Associate Professor of Psychiatry, Columbia
University College of Physicians and Surgeons, New
York, New York

xv

Stephanie N. Mullins-Sweatt, M.A.


Graduate Student, Department of Psychology, University of Kentucky, Lexington, Kentucky
Ziad Nahas, M.D.
Assistant Professor, Department of Psychiatry; Medical Director, Brain Stimulation Laboratory, Center for
Advanced Imaging Research, Medical University of
South Carolina, Charleston, South Carolina
Edmund C. Neuhaus, Ph.D.
Director, Behavioral Health Partial Hospital, McLean
Hospital, Boston, Massachusetts
John S. Ogrodniczuk, Ph.D.
Assistant Professor, Department of Psychiatry, University of British Columbia, Vancouver, British
Columbia, Canada

Wilson McDermut, Ph.D.


Assistant Professor, Department of Psychology,
St. Johns University, Jamaica, New York, and Staff
Psychologist, Albert Ellis Institute, New York

John M. Oldham, M.D., M.S.


Professor and Chairman, Department of Psychiatry
and Behavioral Sciences, Medical University of South
Carolina, Charleston, South Carolina

Pamela G. McGeoch, M.A.


Graduate Faculty, Department of Psychology, The
New School University, New York, New York;
Psychology Intern, Creedmoor Psychiatric Center,
Queens Village, New York

Joel Paris, M.D.


Professor of Psychiatry, McGill University, Montral,
Qubec, Canada

Thomas H. McGlashan, M.D.


Professor of Psychiatry, Department of Psychiatry,
Yale University School of Medicine, New Haven,
Connecticut
Michael J. Meaney, Ph.D.
James McGill Professor of Medicine and Director,
McGill Program for the Study of Behavior, Genes, and
Environment, Douglas Hospital Research Center,
McGill University, Montral, Qubec, Canada
Theodore Millon, Ph.D., D.Sc.
Dean and Scientific Director, Institute for Advanced
Studies in Personology and Psychopathology, Coral
Gables, Florida; Postdoctoral Fellow, Florida International University, Miami, Florida

J. Christopher Perry, M.P.H., M.D.


Professor of Psychiatry, McGill University; Director of
Psychotherapy Research, Institute of Community and
Family Psychiatry, Sir Mortimer B. Davis Jewish General Hospital, Montral, Qubec, Canada; Research
Affiliate, The Austen Riggs Center, Stockbridge, Massachusetts
William E. Piper, Ph.D.
Professor and Head, Division of Behavioural Science;
Director, Psychotherapy Program, Department of Psychiatry, University of British Columbia, Vancouver,
British Columbia, Canada
Christopher J. Pittenger, M.D., Ph.D.
Neuroscience Research Training Program, Department of Psychiatry, Yale University, New Haven,
Connecticut

Chris Molnar, Ph.D.


Postdoctoral Fellow, Brain Stimulation Laboratory,
Center for Advanced Imaging Research, Medical University of South Carolina, Charleston, South Carolina

Christian Schmahl, M.D.


Assistant Medical Director, Department of Psychosomatic Medicine and Psychotherapy, Central Institute
of Mental Health, Mannheim, Germany

Leslie C. Morey, Ph.D.


Professor of Psychology, Department of Psychology,
Texas A&M University, College Station, Texas

Abigail Schlesinger, M.D.


Child Fellow, Western Psychiatric Institute and Clinic,
University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania

xvi

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

M. Tracie Shea, Ph.D.


Associate Professor, Department of Psychiatry and
Human Behavior, Brown University Medical School,
Providence, Rhode Island
G. Pirooz Sholevar, M.D.
Clinical Professor of Psychiatry, Jefferson Medical
College, Thomas Jefferson University, Philadelphia,
Pennsylvania
Larry J. Siever, M.D.
Executive Director, Mental Illness Research, Education and Clinical Center, Bronx Veterans Administration Medical Center, Bronx, New York; Professor of
Psychiatry, Department of Psychiatry, The Mount Sinai School of Medicine, New York, New York
Kenneth R. Silk, M.D.
Professor and Associate Chair, Clinical and Administrative Affairs, University of Michigan Health System,
Ann Arbor, Michigan
Andrew E. Skodol, M.D.
Professor of Clinical Psychiatry, Columbia University
College of Physicians and Surgeons, and Director, Department of Personality Studies, New York State Psychiatric Institute, New York, New York
George W. Smith, M.S.W.
Director, Outpatient Personality Disorder Services,
McLean Hospital, Boston, Massachusetts
Paul H. Soloff, M.D.
Professor of Psychiatry, Western Psychiatric Institute
and Clinic, Pittsburgh, Pennsylvania
David Spiegel, M.D.
Willson Professor and Associate Chair of Psychiatry
and Behavioral Sciences, Department of Psychiatry
and Behavioral Sciences, Stanford University School
of Medicine, Stanford, California; Director, Center for
Integrative Medicine, Stanford Hospital and Clinics
Barbara Stanley, Ph.D.
Lecturer, Department of Psychiatry, Columbia University College of Physicians and Surgeons; Research
Scientist, Department of Neuroscience, New York
State Psychiatric Institute; and Professor, Department
of Psychology, City University of New YorkJohn Jay
College, New York, New York

Michael H. Stone, M.D.


Professor of Clinical Psychiatry, Columbia College of
Physicians and Surgeons, New York, New York
Svenn Torgersen, Ph.D.
Professor, Department of Psychology, University of
Oslo, Blindern, Norway
Peter Tyrer, M.D.
Professor of Community Psychiatry and Head of Department, Department of Psychological Medicine, Imperial College, London, United Kingdom
Louisa M.C. van den Bosch, Ph.D.
Clinical Psychologist and Administrative Executive,
Forensic Psychiatric Hospital, Oldenkotte, Eibergen,
The Netherlands
Roel Verheul, Ph.D.
Professor of Personality Disorders, Viersprong Institute for Studies on Personality Disorders (VISPD),
Center of Psychotherapy De Viersprong, Halsteren,
University of Amsterdam, Department of Clinical
Psychology, Amsterdam, The Netherlands
Drew Westen, Ph.D.
Professor, Department of Psychiatry and Behavioral
Sciences and Department of Psychology, Emory University, Atlanta, Georgia
Thomas A. Widiger, Ph.D.
Professor, Department of Psychology, University of
Kentucky, Lexington, Kentucky
Frank E. Yeomans, M.D.
Clinical Associate Professor of Psychiatry, Department of Psychiatry, Weill Medical College of Cornell
University, New York, New York
Jeffrey Young, Ph.D.
Assistant Professor of Clinical Psychology in Psychiatry, Department of Psychiatry, Columbia University;
Director, Cognitive Therapy Centers of New York &
Connecticut; Director, Schema Therapy Institute, New
York, New York
Mark Zimmerman, M.D.
Associate Professor, Department of Psychiatry and
Human Behavior, Brown University School of Medicine, and Director of Outpatient Psychiatry, Department of Psychiatry, Rhode Island Hospital,
Providence, Rhode Island

Introduction

From as early as the fifth century B.C., it has been recognized that every human being develops an individualized signature pattern of behavior that is reasonably
persistent and predictable throughout life. Hippocrates
proposed that the varieties of human behavior could be
organized into what we might now call prototypes
broad descriptive patterns of behavior characterized by
typical, predominant, easily recognizable features
and that most individuals could be sorted into these
broad categories. Sanguine, melancholic, choleric, and
phlegmatic types of behavior were, in turn, thought to
derive from body humors, such as blood, black bile,
yellow bile, and phlegm, and the predominance of a
given body humor in an individual was thought to correlate with a particular behavior pattern. Although we
now call body humors by different names (neurotransmitters, transcription factors, second messengers), the
ancient principle that fundamental differences in biology correlate with relatively predictable patterns of behavior is strikingly familiar.
In spite of long-standing worldwide interest in personality types, however, remarkably little progress has
been made, until recently, in our understanding of
those severe and persistent patterns of inner experience and behavior that result in enduring emotional
distress and impairment in occupational functioning
and interpersonal relationshipsthe conditions we
now refer to as personality disorders. For decades, it
was widely recognized that some severely disturbed
individuals just seemed to have been born that way,
a view we now know to be true in some cases involving significant genetic loading or risk. In the twentieth
century, however, we became more interested in the
role of the environment during early development in
determining the shape of lasting adult behaviora
view that for a while extended well beyond the realm
of the personality disorders to include most major

mental disorders. We know, of course, that the early life


environment is indeed critically importantfrom
health-promoting, highly nurturing environments to
stressful and neglectful environments from which only
the most resilient emerge unscathed. But we also know
that variable degrees of genetic risk predispose many
of us to become ill in very specific ways, should we unluckily encounter more stress than we can tolerate.
In recent years, we have begun to see an upsurge of
empirical and clinical interest in personality disorders.
Improved standardized diagnostic systems have led to
semistructured research interviews that are being used
not only in studies of clinical populations but also in
community-based studies, to give us, for the first time,
good data about the epidemiology of these disorders.
Personality disorders represent about 12% of the general population, and their public health significance has
been documented by studies showing their extreme social dysfunction and high health care utilization. As
clinical populations are becoming better defined, new
and more rigorous treatment studies are being carried
out, with increasingly promising results. No longer are
personality disorders swept into the hopeless cases
bin. An explosion of knowledge and technology in the
neurosciences has made the formerly black box, the
brain, more and more transparent. Mapping the human genome paved the way for new gene-finding
technologies that are being put to work to tackle complex psychiatric disorders, including the personality
disorders. New transgenic animal models are providing important hints about the genetic loci driving certain behavior types, such as attachment and bonding
behavior. Brain imaging studies are allowing researchers to zero in on malfunctioning areas of the brain in
specific personality disorders.
A great deal of work must still be done. Fundamental questions remain, such as what is the relationship

xvii

xviii

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

between traits of general, or normal, personality functioning and personality psychopathology. Directly related to this issue is the ongoing debate about whether
dimensional or categorical systems best capture the full
scope of personality differences and personality pathology. The extent of impairment associated with personality disorders highlights the significance of gaining
knowledge regarding their longer-term course and increased understanding of factors contributing to variations in course. But there is a strong momentum of interest internationally in these issues, as new research
findings emerge daily to inform the process.
In light of the acceleration of interest and progress
in the field of personality studies and personality disorders, we judged the time to be right to develop a
comprehensive textbook of personality disorders, recognizing that comprehensive coverage of the field
would be a daunting goal and that even newer findings would likely appear by the time the book was
published. However, our attempt has been to assemble as many of the best experts in the field as we could,
to present a thorough and informative survey of what
we now know about the personality disorders. Thus,
this book is organized into several parts: 1) Basic Concepts, 2) Clinical Evaluation, 3) Etiology, 4) Treatment,
5) Special Problems and Populations, and 6) New Developments and Future Directions.

PART I: BASIC CONCEPTS


Basic Concepts, the first part of The American Psychiatric Publishing Textbook of Personality Disorders, might be
thought of as setting the stage for the parts that follow.
In Chapter 1, Oldham presents a brief overview of the
recent history of the personality disorders, along with
a summary look at current controversies and possible
future developments in the field. Heim and Westen, in
the next chapter, review the major theories that have
influenced our thinking about the nature of personality and personality disorders. In Chapter 3, Widiger
and Mullins-Sweatt discuss in depth the arguments
and evidence supporting either categorical models of
personality pathology or dimensional, continuous
models of personality styles and disorders.

PART II: CLINICAL EVALUATION


In the section on clinical evaluation beginning with
Chapter 4, Skodol reviews the defining features of
DSM-IV-TR personality disorders, discusses comple-

mentary approaches to the clinical assessment of a patient with a possible personality disorder, provides
guidance on general problems encountered in the routine clinical evaluation, and describes patterns of Axis
I and Axis II disorder comorbidity. The chapter concludes with a disorder-by-disorder discussion of specific problems in the differential diagnosis of the personality disorders and how the clinician might resolve
them. In Chapter 5, McDermut and Zimmerman review the assessment instruments available for conducting standardized evaluations of personality disorders, including semistructured interviews, other
clinician-administered instruments, and self-report
questionnaires. The instruments covered are those
that measure personality psychopathology according
to the DSM-IV-TR taxonomy, as well as those that
measure alternative concepts of personality and its pathology, such as the Five-Factor Model. Part II concludes with Chapter 6, in which Grilo and McGlashan
provide an overview of the clinical course and outcome of personality disorders, synthesizing the empirical literature on the stability of personality disorder psychopathology.

PART III: ETIOLOGY


The section on etiology of the personality disorders begins with Chapter 7, a presentation by Paris of an integrative perspective on the personality disorders. Paris
reviews the increasingly useful bidirectional stressdiathesis framework, along with its relevance to our
understanding of the dual roles of genes and environment in the etiology of the personality disorders. Torgersen then presents, in Chapter 8, the best data we have
to date on the population-based epidemiology of the
personality disorders. Although there are relatively few
well-designed population-based studies, Torgersen selects eight studies, including his own Norwegian study,
and tabulates prevalence ranges and averages for individual DSM-defined personality disorders as well as for
all personality disorders taken together (showing an
overall average prevalence rate for the personality disorders of over 12%). Of particular interest in these data are
cross-cultural comparisons, suggesting significant cultural differences in the prevalence of selected personality
disorders. The genetic role in the etiology of personality
disorders is summarized by Cloninger in Chapter 9, who
argues that personality styles and disorders are comprised of multiple heritable dimensions, variably expressed, in combination with environmental factors.
Substantial progress has been made in our under-

Introduction

standing of these genetic influences, and new findings


are emerging steadily on the neurobiology of the personality disorders, as reviewed in Chapter 10 by Coccaro
and Siever. Although a great deal more is known about
the neurobiology of some personality disorders (e.g.,
schizotypal personality disorder and borderline personality disorder) than others (e.g., Cluster C personality
disorders), the underlying neurobiological dysfunction
involved in personality disorders characterized by
cognitive symptomatology, impulsivity, and mood dysregulation is becoming increasingly clear.
Understanding the etiology of the personality disorders involves not just cross-sectional genetic and neurobiological analysis; environmental influences shaping
personality must be understood as well. In Chapter 11,
Cohen and Crawford provide a developmental perspective. Although by convention DSM-IV-TR personality
disorders are generally not diagnosed until late adolescence, there is increasing recognition of early patterns of
behavior that are thought to be precursors to certain personality disorders. The challenge to identify true early
precursors of personality disorders, versus the risk of inaccurate labeling of transient symptoms, is central to the
work ahead of us as we focus more and more on prevention strategies. Developmental issues are central to an increasingly persuasive mentalization model of understanding borderline personality disorder, deriving from
basic concepts of attachment theoryreviewed in Chapter 12 by Fonagy and Bateman. In this model, borderline
personality disorder is seen as dysfunction in self-regulation, critically related to interpersonal dynamics. Complementing this model specific to borderline personality
disorder, the authors of Chapter 13, Johnson, Bromley,
and McGeoch, review the relevance of childhood experiences in the development of maladaptive personality
traits. Consistent with the stress-diathesis model presented earlier by Paris in Chapter 7, Johnson and colleagues emphasize not just the importance of stress, but
also the role of protective factors that can offset and even
prevent the development of maladaptive traits in vulnerable individuals. Finally, the section on etiology closes
with Chapter 14, a thoughtful review by Millon and
Grossman of the many sociocultural factors that shape
our behavior, both ordered and disordered.

PART IV: TREATMENT


The treatment section begins with Chapter 15, a discussion of the levels of care available for patients with
personality disorders. Gunderson, Gratz, Neuhaus,
and Smith offer guidelines for determining the appro-

xix

priate intensity of treatment services for individual


patients. Four levels of care are addressed: hospital,
partial hospitalization/day treatment, intensive outpatient, and outpatient.
Chapters 16 through 21 offer a range of outpatient
treatment options that are, for the most part, centered
on interventions within a patient-therapist dyad. Gabbard (Chapter 16) summarizes the salient features of
psychoanalysis as applied to patients with character
pathology, while Yeomans, Clarkin, and Levy offer a review of various psychodynamic psychotherapy approaches in Chapter 17. In the cognitive-behavioral
realm, Chapter 18 by Young and Klosko describes the
latest schema-therapy developments for personality
disorders, and in Chapter 19, Stanley and Brodsky outline the core elements of dialectical behavior therapy,
which includes individual and group interventions,
and is chiefly used to treat parasuicidal behaviors in patients with borderline personality disorder. Patients
with borderline pathology are also the focus of a new
treatment approach based on interpersonal principles,
presented by Markowitz in Chapter 20. Appelbaums
(Chapter 21) synthesis of theories and techniques underpinning supportive psychotherapy provides a fundamental backdrop for many clinicians engaged in the
treatment of personality disorders.
Apart from the realm of individual treatments, there
are other venues for therapeutic interventions. In Chapter 22, Piper and Ogrodniczuk demonstrate the application of group therapy to personality disorders, and the
family is the context for Sholevars work, detailed in
Chapter 23. In addition, Hoffman and Fruzzetti (Chapter 24) suggest various psychoeducational programs
that might benefit personality disorder patients and
their families. Further, Soloff (Chapter 25) takes up the
issue of pharmacotherapy and other somatic treatments, because many patients with personality disorders may benefit by complementing their psychosocial
treatments with medication.
The final three chapters of this section address issues of great importance pertaining to most, if not all,
treatments. Bender (Chapter 26) underscores the necessity of explicitly considering alliance-building
across all treatment modalities, while Gutheil (Chapter 27) cautions practitioners about dynamics that can
lead treaters to boundary violations when working
with certain patients with personality disorders. Finally, as many of these patients with personality disorders are engaged in several modalities with several clinicians at the same time, Schlesinger and Silk, in
Chapter 28, provide recommendations about the best
way of negotiating collaborative treatments.

xx

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

PART V: SPECIAL PROBLEMS AND


POPULATIONS
In recognition of the fact that patients with personality
disorders can be particularly challenging, we have devoted a section of the Textbook to special problems and
populations. Of prime importance is the risk for suicide. In Chapter 29, Links and Kolla provide evidence
on the association of suicidal behavior and personality
disorders, examine modifiable risk factors, and discuss
clinical approaches to the assessment and management
of suicide risk. In Chapter 30, Verheul, van den Bosch,
and Ball focus on pathways to substance abuse in patients with personality disorders, and discuss issues of
differential diagnosis and treatment.
Patients with personality disorders may be not only
a danger to themselves, but also sometimes a danger to
others. Stone, in Chapter 31, discusses aggression and
violence associated with specific personality disorder
types and factors predisposing to violent behavior. The
chapter is illustrated by many clinical vignettes from literature and Stones personal clinical experience. Chapter 32, by Maldonado and Spiegel, is a review of the literature on dissociative states and their relationship to
personality disorder psychopathology. Chapter 33, by
Perry and Bond, presents the theory and measurement
of defense mechanisms relevant to personality disorders, with a discussion of how the management and interpretation of defenses can further psychotherapy.
This chapter also includes many clinical examples of
defenses observed in specific therapeutic interactions.
Gender and culture play important roles in the
evaluation and treatment of personality disorders.
These issues are dissected in Chapters 34 and 35. In
Chapter 34, Morey, Alexander, and Boggs look at gender differences in the prevalence of personality disorders, discuss research bearing on the issue of gender
bias in the diagnosis of personality disorders that may
or may not account for gender distributions, and, finally, describe the interaction of biological and social
factors in determining gender differences in personality traits and behaviors. In Chapter 35, Alarcn discusses the role of culture in the etiology, diagnosis,
and treatment of personality disorders.
As personality disorders have received greater attention from the mental health fields, it has become increasingly apparent that they may be encountered outside of traditional mental health treatment settings,
where they can present special problems in detection
and management. In Chapter 36, Coid describes personality disorders as they are found in prison popula-

tions and the role of personality disorders in determining the risks for the development of career criminals.
Tyrers Chapter 37 on the significance of personality
disorders occurring in the medically ill concludes the
section on special problems and populations.

PART VI: NEW DEVELOPMENTS AND


FUTURE DIRECTIONS
In the final section of The American Psychiatric Publishing
Textbook of Personality Disorders, we have selected a few
areas in which research is intensifying and key findings
are anticipated that will increase our understanding of
the personality disorders. In Chapter 38, Nahas, Molnar, and George review brain imaging studies of patients with personality disorders. Both structural and
functional imaging studies are beginning to shed light
on dysfunctions in the brain in a number of the personality disorders, particularly in schizotypal personality
disorder, borderline personality disorder, and antisocial personality disorder. The very application of basic
research methods to the study of personality disorders
is illustrated not only by brain imaging research but by
the utility of the principles of translational research,
illustrated in Chapter 39 by Bohus and Schmahl, and by
the relevance of animal models for the study of personality disorders, reviewed in Chapter 40 by Meaney.
Finally, in Chapter 41, Etkin, Pittenger, and Kandel present what is currently known about biological changes
in the brain produced by psychotherapy, from the vantage point of psychotherapy as a form of learning. Of
particular relevance, they suggest that neuroimaging
techniques may enable us to identify brain substrates
that are particularly relevant to patients with personality disorders, in order to guide prediction of treatment
outcome.
We are grateful to all of the authors of each chapter
for their careful and thoughtful contributions, and we
hope that we have succeeded in providing a current,
definitive review of the field. We would particularly
like to thank Liz Bednarowicz for her organized and
steadfast administrative support, without which this
volume would not have been possible.
John M. Oldham, M.D., M.S.
Charleston, South Carolina

Andrew E. Skodol, M.D.


New York, New York

Donna S. Bender, Ph.D.


New York, New York

Part I
Basic Concepts

This page intentionally left blank

1
Personality Disorders
Recent History and Future Directions
John M. Oldham, M.D., M.S.

newborn nursery, from cranky to placid. Each individuals temperament remains a key component of
that persons developing personality, to which is
added the shaping and molding influences of family,
caretakers, and environmental experiences. This process is, we now know, bidirectional, so that the inborn behavior of the infant can elicit behavior in parents or caretakers that can, in turn, reinforce infant
behavior: placid, happy babies may elicit warm and
nurturing behaviors; irritable babies may elicit impatient and neglectful behaviors.
However, even-tempered, easy-to-care-for babies
can have bad luck and land in a nonsupportive or
even abusive environment that may set the stage for a
personality disorder, and difficult-to-care-for babies
can have good luck and be protected from future personality pathology by specially talented and attentive
caretakers. Once these highly individualized dynamics have had their main effects and an individual has
reached late adolescence or young adulthood, his or
her personality will usually have been pretty well established. We know that this is not an ironclad rule;

PERSONALITY TYPES
AND PERSONALITY DISORDERS
Charting a historical review of efforts to understand
personality types and the differences among them
would involve exploring centuries of scholarly archives, worldwide, on the varieties of human behavior. For it is human behavior, in the end, that serves as
the most valid measurable and observable benchmark
of personality. In many important ways, we are what
we do. The what of personality is easier to come by
than the why, and each of us has a personality style
that is unique, almost like a fingerprint. At a school reunion, recognition of classmates not seen for decades
derives as much from familiar behavior as from physical appearance.
As to why we behave the way we do, we know
now that a fair amount of the reason relates to our
hardwiring. To varying degrees, heritable temperaments that vary widely from one individual to another determine the amazing range of behavior in the

Sections of this chapter have been modified with permission from Oldham JM, Skodol AE: Charting the Future of Axis II.
Journal of Personality Disorders 14:1729 2000.

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

there are late bloomers, and high-impact life events


can derail or reroute any of us. How much we can
change if we need and want to is variable, but change
is possible. How we define the differences between
personality styles and personality disorders, how the
two relate to each other, what systems best capture the
magnificent variety of nonpathological human behavior, and how we think about and deal with extremes of
behavior that we call personality disorders are all
spelled out in great detail in the chapters of this textbook. In this first chapter, I briefly describe how psychiatrists in the United States have approached the
definition and classification of the personality disorders, building on broader international concepts and
theories of psychopathology.

TWENTIETH-CENTURY CONCEPTS OF
PERSONALITY PSYCHOPATHOLOGY
Personality pathology has been recognized in most
influential systems of classifying psychopathology.
The well-known contributions by European pioneers
of descriptive psychiatry, such as Kraepelin (1904),
Bleuler (1924), Kretschmer (1926), and Schneider
(1923) had an important impact on early twentiethcentury American psychiatry. For the most part, Kraepelin, Bleuler, and Kretschmer described personality
types or temperaments, such as aesthenic, autistic,
schizoid, cyclothymic, or cycloid, that were thought to
be precursors or less extreme forms of psychotic conditions, such as schizophrenia or manic-depressive
illnesssystems that can clearly be seen as forerunners of current Axis I/Axis II spectrum models.
Schneider, on the other hand, described a set of psychopathic personalities that he viewed as separate
disorders co-occurring with other psychiatric disorders. Although these classical systems of descriptive
psychopathology resonate strongly with the framework eventually adopted by the American Psychiatric
Association (APA) and published in its Diagnostic and
Statistical Manual of Mental Disorders (DSM), they were
widely overshadowed in American psychiatry during
the mid-twentieth century by theory-based psychoanalytic concepts stimulated by the work of Sigmund
Freud and his followers.
Freud emphasized the presence of a dynamic unconscious, a realm that, by definition, is mostly unavailable to conscious thought but is a powerful motivator of
human behavior (key ingredients of his topographical
model). His emphasis on a dynamic unconscious was

augmented by his well-known tripartite structural theory, a conflict model serving as the bedrock of his psychosexual theory of pathology (Freud 1926). Freud theorized that certain unconscious sexual wishes or
impulses (id) could threaten to emerge into consciousness (ego), thus colliding wholesale with strict conscience-driven prohibitions (superego) and producing
signal anxiety, precipitating unconscious defense
mechanisms and, when these coping strategies prove
insufficient, leading to frank symptom formation. For
the most part, this system was proposed as an explanation for what were called at the time the symptom neuroses, such as hysterical neurosis or obsessive-compulsive
neurosis. During the 1940s, 1950s, and 1960s, these
ideas became dominant in American psychiatry, followed later by interest in other psychoanalytic principles, such as object relations theory.
Freuds concentration on the symptom neuroses
involved the central notion of anxiety as the engine
that led to defense mechanisms and to symptom formation, and as a critical factor in motivating patients
to work hard in psychoanalysis to face painful realizations and to tolerate stress within the treatment itself
(such as that involved in the transference neurosis).
Less prominently articulated were Freuds notions of
character pathology, but generally character disorders
were seen to represent pre-oedipal pathology. As
such, patients with these conditions were judged less
likely to be motivated to change. Instead of experiencing anxiety related to the potential gratification of an
unacceptable sexual impulse, patients with fixations at the oral-dependent stage, for example, experienced anxiety when not gratifying the impulsein
this case, the need to be fed. Relief of anxiety thus
could be accomplished by some combination of real
and symbolic feedingattention from a parent or parent figure or consumption of alcohol or drugs. Deprivations within the psychoanalytic situation, then
inevitable by its very naturecould lead to patient
flight and interrupted treatment.
In a way, social attitudes mirrored and extended
these beliefs such that although personality pathology was well known, it was often thought to reflect
weakness of character or willfully offensive or socially deviant behavior produced by faulty upbringing, rather than understood as legitimate psychopathology. A good example of this view could be seen
in military psychiatry in the mid-1900s, where those
discharged from active duty for mental illness, with
eligibility for disability and medical benefits, did not
include individuals with character disorders (or alcoholism and substance abuse) because these condi-

Personality Disorders: Recent History and Future Directions

tions were seen as bad behavior and led to administrative, nonmedical separation from the military.
In spite of these common attitudes, clinicians recognized that many patients with significant impairment in social or occupational functioning, or with
significant emotional distress, needed treatment for
psychopathology that did not involve frank psychosis
or other syndromes characterized by discrete, persistent symptom patterns such as major depressive episodes, persistent anxiety, or dementia. General clinical
experience and wisdom guided treatment recommendations for these patients, at least for those who
sought treatment. Patients with paranoid, schizoid, or
antisocial patterns of thinking and behaving often did
not seek treatment. Others, however, often resembled
patients with symptom neuroses and did seek help for
problems ranging from self-destructive behavior to
chronic misery. The most severely and persistently
disabled of these patients were often referred for intensive, psychoanalytically oriented long-term inpatient treatment at treatment centers such as Austen
Riggs, Chestnut Lodge, Menninger Clinic, McLean
Hospital, New York Hospital Westchester Division,
New York State Psychiatric Institute, Sheppard Pratt,
and other long-term inpatient facilities available at the
time. Other patients, able to function outside of a hospital setting and often hard to distinguish from patients with neuroses, were referred for outpatient psychoanalysis or intensive psychoanalytically oriented
psychotherapy. As Gunderson (2001) described, the
fact that many such patients in psychoanalysis regressed and seemed to get worse, rather than showing
improvement in treatment, was one factor that contributed to the emerging concept of borderline personality disorder (BPD), thought initially to be in the border zone between the psychoses and the neuroses.
Patients in this general category included some who
had previously been labeled as having latent schizophrenia (Bleuler 1924), ambulatory schizophrenia
(Zillborg 1941), pseudoneurotic schizophrenia (Hoch
and Polatin 1949), psychotic character (Frosch 1964),
or as-if personality (Deutsch 1942).
These developments coincided with new approaches based on alternative theoretical models that
were emerging within the psychoanalytic framework,
such as the British object relations school. New conceptual frameworks, such as Kernbergs (1975) model
of borderline personality organization or Kohuts
(1971) concept of the central importance of empathic
failure in the histories of narcissistic patients, served
as the basis for an intensive psychodynamic treatment approach for selected patients with personality

disorders. These strategies and others are reviewed in


detail in Chapter 16, Psychoanalysis.

The DSM System


Contrary to assumptions commonly encountered,
personality disorders have been included in every edition of the APAs Diagnostic and Statistical Manual of
Mental Disorders. Largely driven by the need for standardized psychiatric diagnosis in the context of World
War II, the United States War Department in 1943 developed a document labeled Technical Bulletin 203,
representing a psychoanalytically oriented system of
terminology for classifying mental illness precipitated
by stress (Barton 1987). The APA charged its Committee on Nomenclature and Statistics to solicit expert
opinion and to develop a diagnostic manual that
would codify and standardize psychiatric diagnoses.
This diagnostic system became the framework for the
first edition of DSM (DSM-I; American Psychiatric Association 1952). This manual was widely utilized, and
it was subsequently revised on several occasions,
leading to DSM-II (American Psychiatric Association
1968), DSM-III (American Psychiatric Association
1980), DSM-III-R (American Psychiatric Association
1987), DSM-IV (American Psychiatric Association
1994), and DSM-IV-TR (American Psychiatric Association 2000). Figure 11 (Skodol 1997) portrays the ontogeny of diagnostic terms relevant to the personality
disorders from DSM-I through DSM-IV (DSM-IV-TR
involved only text revisions; it used the same diagnostic terms as DSM-IV).
Although not explicit in the narrative text, DSM-I
reflected the general view of personality disorders at
the time, elements of which persist to the present.
Generally, personality disorders were viewed as more
or less permanent patterns of behavior and human interaction that were established by early adulthood and
were unlikely to change throughout the life cycle.
Thorny issues such as how to differentiate personality
disorders from personality styles or traits, which remain actively debated today, were clearly identified at
the time. Personality disorders were contrasted with
the symptom neuroses in a number of ways, particularly that the neuroses were characterized by anxiety
and distress, whereas the personality disorders were
often ego-syntonic and thus not recognized by those
who had them. Even today, we hear descriptions of
some personality disorders as externalizingthat
is, disorders in which the patient disavows any problem but blames all discomfort on the real or perceived
unreasonableness of others. Notions of personality

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

DSM -I (1952)
Personality pattern
disturbance
Inadequate
Paranoid
Cyclothymic
Schizoid

Personality trait
disturbance
Emotionally unstable
Passive- aggressive
dependent type
aggressive type

DSM -II (1968)

DSM -III (1980)


Axis I cyclothymic
disorder

Inadequate
Paranoid
Cyclothymic
Schizoid

Cluster A
Paranoid
Schizoid
Schizotypal

Passive - aggressive

Obsessive - compulsive

Axis I cyclothymic
disorder
Cluster A
Paranoid
Schizoid
Schizotypal

Cluster B

Cluster B
Hysterical

Histrionic
Antisocial
Borderline
Narcissistic

Histrionic
Antisocial
Borderline
Narcissistic
Cluster C

Cluster C
Compulsive

DSM-IV (1994)/
DSM-IV-TR (2000)

Compulsive
Avoidant
Dependent
Passive - aggressive

Obsessive- compulsive
Avoidant
Dependent

Sociopathic personality
disturbance
Antisocial
Dyssocial

Asthenic
Antisocial
Explosive

Axis I intermittent
explosive disorder

Axis I intermittent
explosive disorder
DSM-IV Appendix
Passive -aggressive
Depressive

Indicates that category was discontinued.

Figure 11.

DSM - III - R Appendix*


Self- defeating
Sadistic

Ontogeny of personality disorder classification.

*No changes were made to the personality disorder classification in DSM-III-R except for the inclusion of self-defeating and sadistic personality disorders in Appendix A: Proposed Diagnostic Categories Needing Further Study. These two categories were not included in DSM-IV
or in DSM-IV-TR.
Source. Reprinted with permission from Skodol AE: Classification, Assessment, and Differential Diagnosis of Personality Disorders. Journal of Practical Psychiatry and Behavioral Health 3:261274, 1997.

psychopathology still resonate with concepts such as


those of Reich (1933/1945), who described defensive
character armor as a lifetime protective shield.
In DSM-I, personality disorders were generally
viewed as deficit conditions reflecting partial developmental arrests or distortions in development secondary to inadequate or pathological early caretaking. The
personality disorders were grouped primarily into
personality pattern disturbances, personality trait

disturbances, and sociopathic personality disturbances. Personality pattern disturbances were viewed as
the most entrenched conditions and likely to be recalcitrant to change, even with treatment; these included
inadequate personality, schizoid personality, cyclothymic personality, and paranoid personality. Personality trait disturbances were thought to be less pervasive
and disabling, so that in the absence of stress these patients could function relatively well. If under signifi-

Personality Disorders: Recent History and Future Directions

cant stress, however, patients with emotionally unstable, passive-aggressive, or compulsive personalities
were thought to show emotional distress and deterioration in functioning, and they were variably motivated for and amenable to treatment. The category of
sociopathic personality disturbances reflected what were
generally seen as types of social deviance at the time,
including antisocial reaction, dyssocial reaction, sexual
deviation, and addiction (subcategorized into alcoholism and drug addiction).
The primary stimulus leading to the development
of a new, second edition of DSM was the publication
of the eighth edition of the International Classification of Diseases (World Health Organization 1968)
and the wish of the APA to reconcile its diagnostic terminology with this international system. In the DSM
revision process, an effort was made to move away
from theory-derived diagnoses and to attempt to
reach consensus on the main constellations of personality that were observable, measurable, enduring, and
consistent over time. The earlier view that patients
with personality disorders did not experience emotional distress was discarded, as were the DSM-I subcategories of personality pattern, personality trait,
and sociopathic personality disturbances. One new
personality disorder was added, called asthenic personality disorder, only to be deleted in the next edition of the DSM.
By the mid 1970s, greater emphasis was placed on
increasing the reliability of all diagnoses; whenever
possible, diagnostic criteria that were observable and
measurable were developed to define each diagnosis.
DSM-III, the third edition of the diagnostic manual,
was developed and introduced a multiaxial system.
Disorders classified on Axis I included those generally
seen as episodic, characterized by exacerbations and
remissions, such as psychoses, mood disorders, and
anxiety disorders. Axis II was established to include
the personality disorders as well as mental retardation; both groups were seen as composed of early onset, persistent conditions, but mental retardation was
understood to be biological in origin, in contrast to
the personality disorders, which were generally regarded as psychological in origin. The stated reason
for placing the personality disorders on Axis II was to
ensure that consideration is given to the possible
presence of disorders that are frequently overlooked
when attention is directed to the usually more florid
Axis I disorders (American Psychiatric Association
1980, p. 23). It is generally agreed that the decision to
place the personality disorders on Axis II led to
greater recognition of the personality disorders and

stimulated extensive research and progress in our understanding of these conditions.


As shown in Figure 11, the DSM-II diagnoses of
inadequate personality disorder and asthenic personality disorder were discontinued in DSM-III. The diagnosis of explosive personality disorder was changed
to intermittent explosive disorder, cyclothymic personality disorder was renamed cyclothymic disorder,
and both of these diagnoses were moved to Axis I.
Schizoid personality disorder was felt to be too broad
a category in DSM-II, and it was recrafted into three
personality disorders: schizoid personality disorder,
reflecting loners who are uninterested in close personal relationships; schizotypal personality disorder,
understood to be on the schizophrenia spectrum of
disorders and characterized by eccentric beliefs and
nontraditional behavior; and avoidant personality disorder, typified by self-imposed interpersonal isolation
driven by self-consciousness and anxiety. Two new
personality disorder diagnoses were added in DSMIII: BPD and narcissistic personality disorder. In contrast to initial notions that patients called borderline
were on the border between the psychoses and the
neuroses, the criteria defining BPD in DSM-III emphasized emotional dysregulation, unstable interpersonal relationships, and loss of impulse control more
than cognitive distortions and marginal reality testing,
which were more characteristic of schizotypal personality disorder. Among many scholars whose work
greatly influenced and shaped our understanding of
borderline pathology were Kernberg (1975) and Gunderson (1984, 2001). Although concepts of narcissism
had been described by Freud, Reich, and others, the
essence of the current views of narcissistic personality
disorder emerged from the work of Millon (1969), Kohut (1971), and Kernberg (1975).
DSM-III-R was published in 1987 after an intensive
process to revise DSM-III involving widely solicited
input from researchers and clinicians and following
similar principles to those articulated in DSM-III, such
as assuring reliable diagnostic categories that were
clinically useful and consistent with research findings
and thus minimizing reliance on theory. Efforts were
made for diagnoses to be descriptive and to require
a minimum of inference, although the introductory
text of DSM-III-R acknowledged that for some disorders, particularly the Personality Disorders, the criteria require much more inference on the part of the
observer (American Psychiatric Association 1987,
p. xxiii). No changes were made in DSM-III-R diagnostic categories of personality disorders, although
some adjustments were made in certain criteria sets,

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

for example, making them uniformly polythetic instead of defining some personality disorders with
monothetic criteria sets (e.g., dependent personality
disorder) and others with polythetic criteria sets (e.g.,
borderline personality disorder). In addition, two personality disorders were included in DSM-III-R in Appendix A (Proposed Diagnostic Categories Needing
Further Study)self-defeating personality disorder
and sadistic personality disorderbased on prior
clinical recommendations to the DSM-III-R personality disorder subcommittee. These diagnoses were considered provisional, pending further review and research.
DSM-IV was derived after an extensive process of
literature review, data analysis, field trials, and feedback from the profession. Because of the increase in research stimulated by the criteria-based multiaxial system of DSM-III, a substantial body of evidence existed
to guide the DSM-IV process. As a result, the threshold for approval of revisions for DSM-IV was higher
than that used in DSM-III or DSM-III-R. DSM-IV introduced, for the first time, a set of general diagnostic criteria for any personality disorder (Table 11), underscoring qualities such as early onset, long duration,
inflexibility, and pervasiveness. Diagnostic categories
and dimensional organization of the personality disorders into clusters remained the same in DSM-IV as
in DSM-III-R, with the exception of the relocation of
passive-aggressive personality disorder from the official diagnostic list to Appendix B (Criteria Sets and
Axes Provided for Further Study). Passive-aggressive personality disorder, as defined by DSM-III and
DSM-III-R, was thought to be too unidimensional and
generic; it was tentatively retitled negativistic personality disorder, and the criteria were revised. In addition, the two provisional Axis II diagnoses in DSMIII-R, self-defeating personality disorder and sadistic
personality disorder, were dropped because of insufficient research data and clinical consensus to support
their retention. One other personality disorder was
proposed and added to Appendix B: depressive personality disorder. Although substantially controversial, this provisional diagnosis was proposed as a pessimistic cognitive style; its validity and its distinction
from passive-aggressive personality disorder on Axis
II or dysthymic disorder on Axis I, however, remain to
be established.
DSM-IV-TR, published in 2000, did not change the
diagnostic terms or criteria of DSM-IV. The intent of
DSM-IV-TR was to revise the descriptive, narrative
text accompanying each diagnosis where it seemed indicated and to update the information provided. Only

Table 11.

General diagnostic criteria for a


personality disorder

A. An enduring pattern of inner experience and behavior


that deviates markedly from the expectations of the
individual's culture. This pattern is manifested in two
(or more) of the following areas:
(1) cognition (i.e., ways of perceiving and interpreting
self, other people, and events)
(2) affectivity (i.e., the range, intensity, lability, and
appropriateness of emotional response)
(3) interpersonal functioning
(4) impulse control
B.

The enduring pattern is inflexible and pervasive across


a broad range of personal and social situations.

C.

The enduring pattern leads to clinically significant


distress or impairment in social, occupational, or other
important areas of functioning.

D. The pattern is stable and of long duration, and its onset


can be traced back at least to adolescence or early
adulthood.
E.

The enduring pattern is not better accounted for as a


manifestation or consequence of another mental
disorder.

F.

The enduring pattern is not due to the direct


physiological effects of a substance (e.g., a drug of
abuse, a medication) or a general medical condition
(e.g., head trauma).

Source. Reprinted with permission from American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th
Edition, Text Revision. Washington, DC, American Psychiatric Association, 2000.

minimal revisions were made in the text material accompanying the personality disorders.

Current Controversies and Future Directions


There is a general consensus, at least in the United
States, that the placement of the personality disorders
on Axis II has stimulated research and focused clinical
and educational attention on these disabling conditions. However, there is growing debate about the
continued appropriateness of maintaining the personality disorders on a separate axis in future editions of
the diagnostic manual and about whether a dimensional or a categorical system of classification is preferable. As new knowledge has rapidly accumulated
about the personality disorders, these controversies
take their places among many ongoing constructive
dialogues, such as the relationship of normal personality to personality disorder, the pros and cons of
polythetic criteria sets, how to determine the appropriate number of criteria (i.e., threshold) required for
each diagnosis, which personality disorder categories

Personality Disorders: Recent History and Future Directions

have construct validity, which dimensions best cover


the scope of normal and abnormal personality, and
others. Many of these discussions overlap with and inform each other, yet a central issue under scrutiny is
whether or not to maintain a separate diagnostic axis
for the personality disorders. I briefly review some ongoing challenges and debates in the following sections, all of which are examined in greater detail in the
subsequent chapters in this volume.

Dimensional or Categorical?
Much of the literature poses the question of a dimensional or categorical system as a debate or competition,
as though one must choose sides in a dimensional
versus categorical Super Bowl. Blashfield and McElroy (1995) provided helpful clarification about our terminology, pointing out that a categorical model is a
more complex, elaborated version of a dimensional
model (p. 409). They noted that the DSM-IV system
includes 10 categories grouped into three dimensions
(called clusters), and they clarified, as did Livesley et
al. (1994), Clark (1995), Widiger (1993) and others
(Gunderson et al. 1991; Livesley 1998), that dimensional structure implies continuity whereas categorical
structure implies discontinuity. For example, being
pregnant is a categorical concept (either one is pregnant or one is not, even though we speak of how far
along one is), whereas being tall or short might better
be conceptualized dimensionally, because there is no
exact definition of either, notions of tallness or shortness may vary among different cultures, and all gradations of height exist along a continuum.
We know, of course, that the DSM system is referred to as categorical and is contrasted to any number of systems referred to as dimensional, such as the
interpersonal circumplex (Benjamin 1993; Kiesler
1983; Wiggins 1982), the three-factor model (Eysenck
and Eysenck 1975), several four-factor models (Clark
et al. 1996; Livesley et al. 1993, 1998; Watson et al. 1994;
Widiger 1998), the five-factor model (Costa and McCrae 1992), and the seven-factor model (Cloninger et
al. 1993). How fundamental is the difference between
the two types of systems? Livesley et al. (1994) went so
far as to say that DSM-III-R categorical diagnoses are
based on cutting scores. Individuals who meet more
than a threshold number of criteria are believed to be
qualitatively different from those who meet fewer criteria (p. 8). They added that [a]lthough many of the
features of the DSM-III-R and DSM-IV. ..personality
disorder diagnoses are not substantively different
from features of normal personality, they are used to
define discontinuous categories (p. 8). Although this

concept of discontinuity is implied by a categorical


system, clinicians do not necessarily think in such dichotomous terms. Thresholds defining disease categories, such as hypertension, are in fact somewhat arbitrary, as is certainly the case with the personality
disorders. In addition, the polythetic criteria sets for
the DSM-IV personality disorders contain an element
of dimensionality, because one can just meet the
threshold or can have all of the criteria (and thus presumably a more extreme version of the disorder). Widiger (1991a, 1993) and Widiger and Sanderson (1995)
suggested that this inherent dimensionality in our existing system could be usefully operationalized by
stratifying each personality disorder into subcategories of absent, traits, subthreshold, threshold, moderate, and extreme according to the number of criteria
met. Certainly if an individual is one criterion short of
being diagnosed with a personality disorder, clinicians do not necessarily assume that there is no element of the disorder present; instead, prudent clinicians would understand that features of the disorder
need to be recognized if present and may need attention. Nonetheless, a prudent, thoughtful clinician is
one thing, and a busy, pressured clinician hustling to
get paperwork finished may be another; there would
be a natural tendency to think categoricallythat is, to
decide what disorders the patient officially has and
to disregard all else. In fact, studies of clinical practice
patterns reveal that clinicians generally assign only
one Axis II diagnosis (Westen 1997), whereas systematic studies of clinical populations utilizing semistructured interviews generally reveal multiple Axis II diagnoses and significant traits in individuals who have
pathology on Axis II (Oldham et al. 1992; Shedler and
Westen 2004; Skodol et al. 1988; Widiger et al. 1991).

Definition of a Personality Disorder


As mentioned previously, DSM-IV introduced general criteria defining personality disorders that emphasize the early onset; the primary, enduring and
cross-situational nature of the pathology; and the
presence of emotional distress or impairment in social
or occupational functioning. Although this effort to
specify the generic components of all personality disorders has been helpful, the definition is relatively
nonspecific and could apply to many Axis I disorders
as well, such as dysthymia or even schizophrenia. In
fact, DSM-IV-TR states that it
may be particularly difficult (and not particularly
useful) to distinguish Personality Disorders from
those Axis I disorders (e.g., Dysthymic Disorder)
that have an early onset and a chronic, relatively sta-

10

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

ble course. Some Personality Disorders may have a


spectrum relationship to particular Axis I conditions (e.g., Schizotypal Personality Disorder with
Schizophrenia; Avoidant Personality Disorder with
Social Phobia) based on phenomenological or biological similarities or familial aggregation. (American Psychiatric Association 2000, p. 688)

Livesley (1998) and Livesley and Jang (2000) proposed that the two key ingredients of a revised definition for personality disorder might be chronic interpersonal difficulties and problems with a sense of self,
notions consistent with Kernbergs umbrella concept of
borderline personality organization (Kernberg 1975)
that encompasses many of the DSM-IV personality disorder categories and also consistent with earlier concepts of personality pathology (Schneider 1923). Livesley (1998) proposed a working definition for personality
disorder as a tripartite failure involving 3 separate but
interrelated realms of functioning: self-system, familial
or kinship relationships, and societal or group relationships (p. 141). This proposed revision was suggested
as one that could more readily be translated into reliable measures and as one that derives from an understanding of the functions of normal personality. Although this definition conceptually links personality
pathology with normal personality traits and emphasizes dimensional continuity, how readily measurable a
failure in a self-system would be seems unclear. More
importantly, this proposed definition could be applied
to major Axis I conditions such as schizophrenia, unless
one added the third criterion for borderline personality
organization described by Kernberg (1975): maintenance of reality testing.
Whether the current generic personality disorder
definition is retained or a new one such as that just described were to be adopted, there would still be a need
for specified types of personality disordersretaining
or modifying the existing categories or replacing them
with selected dimensions. In either case, criteria defining the types would be needed. Problems with the current criteria include the hodgepodge mixture of traits
and behavioral measures, a confusion that has been
criticized (Livesley and Jackson 1992; Widiger 1991a).
Widiger (1991a) described the problems in the DSM
system that resulted from the unsuccessful efforts of
the DSM authors to devise criteria sets that define each
personality disorder and that provide measures with
which to diagnose each disorder at the same time. Initial attempts by the DSM-III committee to include only
measurable and observable (i.e., behavioral) criteria
were most evident in the much-criticized criteria set
for antisocial personality disorder (seen as a checklist

for criminal behavior that omitted lack of remorse


[later added in DSM-III-R], a fundamental defining
feature of psychopathy), yet not so evident in other
cases such as narcissistic personality disorder (which
did include lack of empathy as one type of disturbance in interpersonal relationshipsa defining feature of the concept of narcissism rather than a readily
assessed or measured behavior).
Widiger (1991a) suggested that two criteria sets
might be devised, one to define a disorder and a different one to diagnose it, but he admitted that sufficiently comprehensive behavioral criteria sets would
be too lengthy to be practical. Livesley and Jackson
(1992) also suggested developing a definitional system based on expert opinion and complemented by a
set of diagnostic exemplars, each of which should be
direct, noncomplex, and relevant to only one trait of a
diagnosis and only one diagnosis. Although perhaps
this model represents a laudable goal, it would be a
daunting challenge to identify such specific behavioral criteria (exemplars). For example, a simple, direct, measurable behavior such as spending most of
ones time alone could reflect anxiety, depression, low
self-esteem, lack of self-confidence, schizoid disconnectedness, or paranoid suspiciousness. Finally, even
if one succeeded in developing a reasonably representative set of behavioral criteria considered diagnostic
of a personality disorder, it is unlikely, as a number of
authors have pointed out (Gunderson 1987; Widiger
1991a), that such a set would be optimal in all situations, because personality pathology is often activated
or intensified by circumstance, such as loss of a job or
of a meaningful relationship. In the ongoing findings
of the Collaborative Longitudinal Study of Personality
Disorders (Grilo et al. 2004; Shea et al. 2002), this problem has become evident, because stability of diagnosis
must rely on sustained pathology above the DSM-IVTR diagnostic threshold, and substantial percentages
of patients show fluctuation over time, sometimes being above and sometimes below the diagnostic threshold. These data support an argument for a more flexible dimensional component to our diagnostic system,
perhaps along the lines of Widiger s stratification
scheme (Widiger 1991a, 1993; Widiger and Sanderson
1995).
One suggested way to better capture the essence of
each personality disorder is to define the classic
casethat is, the prototype. Livesley (1986, 1987) utilized DSM-III categories and reported that clinicians
could reliably agree on prototypical traits and behaviors of the personality disorders. Widiger (1991a) cautioned, however, that such prototypes might not apply

Personality Disorders: Recent History and Future Directions

to most cases seen in clinical practice and thus might


be of little utility. Although it does seem clear that clinicians can prioritize the criteria of each Axis II diagnosis when asked to list, in order of importance, the
criteria they believe to be most representative of the
disorder, different information may be obtained when
clinicians are asked different questions. Westen and
Arkowitz-Westen (1998) reported the results of a survey of clinicians who were asked if they were treating
patients with personality pathology who could not be
diagnosed on Axis II. They found, in a survey of clinicians, that over 60% of patients reported to have personality pathology for which treatment was indicated
were currently undiagnosable on Axis II. The results
suggested that much of the personality pathology clinicians see and treat in practice may not be captured
by Axis II of DSM-IV (p. 1767). Westen and Shedler
(1999a, 1999b) devised a method based on a Q-score
system to develop clinician-derived prototypes. They
presented seven Q-factors (dysphoric, antisocial-psychopathic, schizoid, paranoid, obsessional, histrionic,
and narcissistic), the psychological features of which,
they proposed, represent coherent, meaningful clinical syndromes. In later work, Shedler and Westen
(2004) again proposed a prototype matching model for
diagnosing personality disorders in the context of concerns about the narrowness of the DSM criteria sets
and the resulting extensive overlap among some diagnostic categories.

Reliability and Validity


Many authors have discussed the continuing questions
of reliability and validity, which inevitably must be
considered together (Clark et al. 1997; Lenzenweger
and Clarkin 1996; Livesley 1998; Perry 1990). Debates
continue regarding the most reliable ways to assess the
Axis II categories. In clinical research, semistructured
interviews have been developed, such as the International Personality Disorder Examination (Loranger
1999), the Structured Interview for DSM-IV Personality
Disorders (Pfohl et al. 1997), and the Structured Clinical
Interview for DSM-IV (First et al. 1997; see Chapter 4,
Manifestations, Clinical Diagnosis, and Comorbidity, and Chapter 5, Assessment Instruments and
Standardized Evaluation). These interviews are called
semistructured because they are administered by a clinician rather than an untrained technician so that the
clinician can probe and explore areas of confusion or
inconsistency and can employ clinical judgment in
making ratings. These methods involve at least two
data sources, the clinician and the patient, and some require input from collateral informants.

11

Studies have repeatedly shown that good interrater reliability can be achieved for most Axis II semistructured interviews, but inter-interview agreement
is consistently poor (Oldham et al. 1992; Perry 1992;
Pilkonis et al. 1991; Skodol et al. 1988). This inability to
obtain the same data from the same patient with different interview instruments may indeed relate to differences in interview construction, but it may also reflect underlying questions about the construct validity
of the diagnostic categories themselves (Livesley 1998;
Perry 1990). Overlapping criteria in many of the categories diminish the points of rarity (Kendell 1975;
Livesley et al. 1994) or discontinuity between categories. Systematic studies reveal high levels of comorbidity within Axis II itself, suggesting that the various
categories may not be independent, valid constructs.
Spitzer (1983) proposed a LEAD (longitudinal expert
evaluation using all data) standard, but operationalization of this standard in clinical research or in efficient clinical care is formidable.

Future Directions
Where do we go from here? Livesley (1998) contended
that [w]hatever advantages accrued from forcing clinicians to consider personality during the diagnostic
process by placing personality disorders on a separate
axis have been realized (p. 139). The problems and
concerns about the justification of maintaining the
personality disorders on Axis II have been discussed
at length (Krueger and Tackett 2003; Millon 2000; Shea
and Yen 2003; Widiger 2003), particularly as potential
changes that might be incorporated into DSM-V are
anticipated. What are the suggestions for change and
how feasible are they?

Move Personality Disorders to Axis I


In the context of addressing the lack of clear differentiation between Axis I and Axis II (Pfohl 1999), Widiger
and Shea (1991) suggested that some Axis II disorders
could be shifted to Axis I, and vice versa. A variation of
this suggestion would be to move some of the personality disorders to Axis I but to retain eachs label as a
personality disorder (or code them on both Axes I and
II). This acknowledges that the Axes I and II boundary
is fluid, at times with no real distinction (p. 402).
Livesley et al. (1994) broadened this suggestion, stating
that [b]ecause personality disorder does not appear to
be substantially different in kind from other mental
disorders, we would prefer to classify personality disorder on Axis I and to use a separate axis (perhaps
Axis II) to code personality traits (p. 14). Arguments

12

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

are increasingly persuasive that Axis II disorders, as


currently defined in DSM-IV-TR, are not fundamentally distinct from Axis I disorders. Nonetheless, there
might still be plausible reasons to maintain the personality disorders on Axis II, in the context of significant
revisions.

Replace the Current Axis II Categorical System


With a Dimensional System
Frances (1993) stated that [s]omeday (perhaps in time
for the fifth edition of [DSM]), we will almost certainly
be applying a dimensional model of personality diagnosis (p. 110). The overwhelming majority of opinion
in the literature on this subject favors the adoption of
some type of dimensional approach (Clark et al. 1996;
Cloninger et al. 1993; Costa and Widiger 1994; First et
al. 2002; Frances 1993; Livesley 1998; Livesley and Jackson 1992; Livesley et al. 1993, 1998; Tellegen 1993; Watson et al. 1994; Westen and Shedler 1999a, 1999b; Widiger 1991a, 1991b, 1992, 1993, 1998; Widiger and Shea
1991). There is evidence supporting the dimensional
view that personality psychopathology represents a
crescendo on the end of a continuous scale defining
personality traits (the hypertension model) (Livesley
et al. 1993, 1998). Conceptualized, then, as exaggerations of normal functioning (intense, extreme, hence
maladaptive personality traits), the challenge to those
creating the diagnostic manual is to develop a scheme
that portrays this dimensional continuity and includes
normal personality types or traits. Advocates of the categorical system contend that such a change would be
too discrepant from traditional medical and clinical tradition and that the categorical system, admittedly a
somewhat artificial convention, should be maintained.

Emphasize Level of Functioning


A number of authors emphasize the importance of
level of functioning in the classification of personality
and personality disorders (Gunderson et al. 1991;
Kernberg 1975; Livesley et al. 1994; Skodol et al. 2002;
Tyrer 1995; Westen and Arkowitz-Westen 1998). Kernbergs (1975) concept of borderline personality organization implies such a hierarchy, distinguishing three
broad categories of intrapsychic structure (neurotic,
borderline, and psychotic) that roughly correlate with
decreasingly successful functioning. Gunderson et al.
(1991) broadened this concept, portraying individuals
with all personality disorders in an intermediate level
between higher-functioning neurotic patients and
lower-functioning psychotic patients. Such schemes

represent, in effect, dimensions of severity (e.g., mild,


moderate, and severe) into which all mental illnesses
could, theoretically, be sorted. In contrast, DSM-IV included impairment in social or occupational functioning as one of the defining criteria for personality disorders, which could then be evaluated utilizing Axis V,
the Global Assessment of Functioning (GAF) Scale.
Skodol et al. (1988) and Goldman et al. (1992) criticized
the use of the GAF Scale because it confounds impairment in social and occupational functioning with
symptom levels. Westen and Arkowitz-Westen (1998)
argued in favor of a functional assessment of personality, representing a case-formulation approach. They
argued that instead of asking diagnostic questions
such as Does the patient cross the threshold for a personality disorder? or How low is the patient on the
trait of agreeableness? a functional assessment would
ask, Under what circumstances are which dysfunctional cognitive, affective, motivational, and behavioral patterns likely to occur? Although approaches
such as these are appealing, they represent a plea to return to the time-honored tradition of careful clinical assessment and formulation; how effectively such systems could be standardized for research purposes or
for clinical use is not clear.

Retain Personality Disorders on Axis II but Collapse


and Stratify the Current Categories
One possible modification of the current system
would be to retain the categorical system but specify
that no patient should be given more than two comorbid personality disorder diagnoses using the existing
categories (Oldham and Skodol 2000; Oldham et al.
1992). In such a model, when three or more personality disorders are determined to be present (above
threshold) in any given patient, a single diagnosis
could be utilized (e.g., extensive personality disorder). Widiger and Sanderson (1995) noted that this
suggestion would eliminate the conceptual and clinical oddity of diagnosing a patient with three, four, or
more purportedly comorbid and distinct personality
disorders (p. 445). They also noted, however, that it
would fail to address the presence of clinically significant traits that are below the diagnostic threshold.
This concern could be addressed in the following way:
for patients with more than two comorbid personality
disorder diagnoses, one could diagnose extensive
personality disorder, characterized by (a, b, c) components (above-threshold categories) and (x, y, z) features (clinically significant traits). The determination
of which below-threshold traits are clinically signifi-

Personality Disorders: Recent History and Future Directions

cant could be either a matter of judgment by the clinician or based on a designated number of criteria met,
as proposed by Widiger (1991a, 1993) and Widiger
and Sanderson (1995).
Although the nature of personality disorders is
not, after all, fundamentally distinct from that of many
disorders on Axis Ihence conceptual consistency
might better be approached by relocating them on
Axis Ia preferable model for DSM-V might be to reconfigure and retain the personality disorders on
Axis II. The primary justification for maintaining the
personality disorders on a separate axis would be to
allow the inclusion of trait assessment in DSM, a manual dedicated to the diagnosis of psychopathology.
Eventually (perhaps beyond DSM-V), continuous concepts could be developed that encompass normal personality styles, personality disorder traits, and personality disorders themselves. Clinicians could evaluate
potentially clinically significant traits within a dimensional and categorical system. (As Tellegen [1993]
stated, [t]he terms dimensional and categorical are
sometimes contrasted as if standing for mutually exclusive alternatives. In reality, valid dimensional and
categorical distinctions exist side by side, both among
indicators and among latent variables [p. 123].) It
would then be possible, by retaining existing or revised personality disorder categories, to stratify them
in a more systematic way, such as that described by
Widiger (1991a, 1993) and Widiger and Sanderson
(1995).
Such a scheme could be readily charted and displayed on a graph like that of the Minnesota Multiphasic Personality Inventory, further conveying the
integration of its categorical and dimensional aspects.
In this proposal, dimensional traits are pathologydefined because they represent the presence of some
of the criteria of the disorders. A more ambitious proposal would be to develop criteria for normal personality types that correspond to their extreme forms
that is, the disorders. Such a DSM-IVbased system
has been described (Oldham and Morris 1995), but the
criteria for normal personality types (or others that
could be developed) would need to be validated.
Finally, a decision to maintain the personality disorders on Axis IIbut to introduce a stratification system
such as that described abovewould not require retention of the exact categories presently included in DSMIV. An empirically based set of diagnoses such as the
prototypes described by Westen and Shedler (1999a,
1999b) could be adopted. This set of categories was developed based on clinician opinion and is based on prototypes derived from clinical constructs closely related

13

to DSM-IV. As a result, these categories would be quite


familiar to the clinical world and could be readily
accepted. Although a broader revision could be attempted that could include normal personality types
and that might incorporate a well-researched dimensional approach such as the five-factor model, such an
undertaking might still be premature for DSM-V.

CONCLUSIONS
This brief review of recent notions of personality pathology serves as a window on the rapid progress in
our field and in our understanding of psychiatric disorders. Increasingly, a stress/diathesis framework
seems applicable in medicine in general as a unifying
model of illnessa model that can easily encompass
the personality disorders (Paris 1999). Variable genetic
vulnerabilities predispose us all to potential future illness that may or may not develop depending on the
balance of specific stressors and protective factors.
The personality disorders represent maladaptive
exaggerations of nonpathological personality styles
resulting from predisposing temperaments combined
with stressful circumstances. Neurobiology can be altered in at least some Axis II disorders, as it can be in
Axis I disorders. Our challenge for the future is to recognize that not all personality disorders are alike, nor
are personality disorders fundamentally different
from many other psychiatric disorders. What may be
somewhat unique to the personality disorders is their
correlation and continuity with normal functioning,
which could be an important consideration in future
revisions of our diagnostic system. As we learn more
about the etiologies and pathology of the personality
disorders, it will no longer be necessary, or even desirable, to limit our diagnostic schemes to atheoretical,
descriptive phenomena, and we can look forward to
an enriched understanding of these disorders.

REFERENCES
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders. Washington, DC, American Psychiatric Association, 1952
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 2nd Edition. Washington,
DC, American Psychiatric Association, 1968
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition. Washington,
DC, American Psychiatric Association, 1980

14

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

American Psychiatric Association: Diagnostic and Statistical


Manual of Mental Disorders, 3rd Edition, Revised. Washington, DC, American Psychiatric Association, 1987
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition. Washington,
DC, American Psychiatric Association, 1994
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association, 2000
Barton WE: The History and Influence of the American Psychiatric Association. Washington, DC, American Psychiatric Press, 1987
Benjamin LS: Interpersonal Diagnosis and Treatment of Personality Disorders. New York, Guilford, 1993
Blashfield RK, McElroy RA: Confusions in the terminology
used for classificatory models, in The DSM-IV Personality Disorders. Edited by Livesley WJ. New York, Guilford, 1995, pp 407416
Bleuler E: Textbook of Psychiatry (English translation). New
York, Macmillan, 1924
Clark LA: The challenge of alternative perspectives in classification: a discussion of basic issues, in The DSM-IV
Personality Disorders. Edited by Livesley WJ. New
York, Guilford, 1995, pp 482496
Clark LA, Livesley WJ, Schroeder ML, et al: Convergence of
two systems for assessing specific traits of personality
disorder. Psychol Assess 8:294303, 1996
Clark LA, Livesley WJ, Morey L: Special feature: personality
disorder assessment: the challenge of construct validity.
J Personal Disord 11:205231, 1997
Cloninger CR, Svrakic DM, Przybeck TR: A psychobiological model of temperament and character. Arch Gen Psychiatry 50:975990, 1993
Costa PT, McCrae RR: The five-factor model of personality
and its relevance to personality disorders. J Personal
Disord 6:343359, 1992
Costa PT, Widiger TA: Personality Disorders and the FiveFactor Model of Personality. Washington, DC, American Psychological Association, 1994
Deutsch H: Some forms of emotional disturbance and their relationship to schizophrenia. Psychoanal Q 11:301321, 1942
Eysenck HJ, Eysenck SBG: Manual of the Eysenck Personality Questionnaire. San Diego, CA, Educational and Industrial Testing Service, 1975
First M, Gibbon M, Spitzer RL, et al: Users Guide for the
Structured Clinical Interview for DSM-IV Axis II Personality Disorders. Washington, DC, American Psychiatric Press, 1997
First MB, Bell CC, Cuthbert B, et al: Personality disorders
and relational disorders: a research agenda for addressing crucial gaps in DSM, in A Research Agenda for
DSM-V. Edited by Kupfer DJ, First MB, Regier DA.
Washington, DC, American Psychiatric Association,
2002, pp 123199
Frances A: Dimensional diagnosis of personality: not
whether, but when and which (commentary). Psychological Inquiry 4:110111, 1993

Freud S: Inhibitions, Symptoms and Anxiety, Standard Edition, London, Hogarth Press, 1926
Frosch J: The psychotic character: clinical psychiatric considerations. Psychiatr Q 38:8196, 1964
Goldman HH, Skodol AE, Lave TR: Revising axis V for
DSM-IV: a review of measures of social functioning. Am
J Psychiatry 149:11481156, 1992
Grilo CM, Shea MT, Sanislow CA et al: Two-year stability
and change in schizotypal, borderline, avoidant, and
obsessive-compulsive personality disorders. J Consult
Clin Psychol 72:767775, 2004
Gunderson JG: Borderline Personality Disorder. Washington, DC, American Psychiatric Press, 1984
Gunderson JG: Interfaces between psychoanalytic and empirical studies of borderline personality, in The Borderline Patient, Vol 1. Edited by Grotstein J, Solomon M,
Lang J. Hillsdale, NJ, The Analytic Press, 1987
Gunderson JG: Borderline Personality Disorder: A Clinical
Guide. Washington, DC, American Psychiatric Publishing, 2001
Gunderson JG, Links PS, Reich JH: Competing models of
personality disorders. J Personal Disord 5:6068, 1991
Hoch PH, Polatin P: Pseudoneurotic forms of schizophrenia.
Psychiatr Q 23:248276, 1949
Kendell RE: The Role of Diagnosis in Psychiatry. Oxford, England, Basil Blackwell, 1975
Kernberg O: Borderline Conditions and Pathological Narcissism. New York, Jason Aronson, 1975
Kiesler DJ: The 1982 interpersonal circle: a taxonomy for
complementarity in human transactions. Psychol Rev
90:185214, 1983
Kohut H: The Analysis of the Self. New York, International
Universities Press, 1971
Kraepelin E: Lectures on Clinical Psychiatry (English translation). New York, Wood Press, 1904
Kretschmer E: Hysteria (English translation). New York,
Nervous and Mental Disease Publishers, 1926
Krueger RF, Tackett MA: Personality and psychopathology:
working toward the bigger picture. J Personal Disord
17:109128, 2003
Lenzenweger MF, Clarkin JF: The personality disorders: history, classification, and research issues, in Major Theories of Personality Disorder. Edited by Clarkin JF, Lenzenweger MF. New York, Guilford, 1996
Livesley WJ: Trait and behavioral prototypes of personality
disorder. Am J Psychiatry 143:728732, 1986
Livesley WJ: A systematic approach to the delineation of
personality disorders. Am J Psychiatry 144:772777,
1987
Livesley WJ: Suggestions for a framework for an empirically
based classification of personality disorder. Can J Psychiatry 43:137147, 1998
Livesley WJ, Jackson DN: Guidelines for developing, evaluating, and revising the classification of personality disorders. J Nerv Ment Disease 180:609618, 1992

Personality Disorders: Recent History and Future Directions

Livesley WJ, Jang KL: Toward an empirically based classification of personality disorder. J Personal Disord 14:137
151, 2000
Livesley WJ, Jang KL, Jackson DN, et al: Genetic and environmental contributions to dimensions of personality
disorder. Am J Psychiatry 150:18261831, 1993
Livesley WJ, Schroeder ML, Jackson DN, et al: Categorical
distinctions in the study of personality disorder: implications for classification. J Abnorm Psychol 103:617, 1994
Livesley WJ, Jang KL, Vernon PA: Phenotypic and genetic
structure of traits delineating personality disorder. Arch
Gen Psychiatry 55:941948, 1998
Loranger AW: International Personality Disorder Examination (IPDE). Odessa, FL, Psychological Assessment Resources, 1999
Millon T: Modern Psychopathology. Philadelphia, PA, WB
Saunders, 1969
Millon T: Reflections on the future of DSM Axis II. J Personal
Disord 14:3041, 2000
Oldham JM, Morris LB: The New Personality Self-Portrait,
New York, Bantam, 1995
Oldham JM, Skodol AE: Charting the future of Axis II. J Personal Disord 14:1729, 2000
Oldham JM, Skodol AE, Kellman HD, et al: Diagnosis of
DSM-III-R personality disorders by two structured interviews: patterns of comorbidity. Am J Psychiatry 149:213
220, 1992
Paris J: Nature and Nurture in Psychiatry: A PredispositionStress Model of Mental Disorders. Washington, DC,
American Psychiatric Press, 1999
Perry JC: Challenges in validating personality disorders: beyond description. J Personal Disord 4:273289, 1990
Perry JC: Problems and considerations in the valid assessment of personality disorders. Am J Psychiatry 149:1645
1653, 1992
Pfohl B: Axis I and Axis II: comorbidity or confusion? in Personality and Pathology. Edited by Cloninger CR. Washington, DC, American Psychiatric Press, 1999, pp 8398
Pfohl B, Blum N, Zimmerman M: Structured Interview for
DSM-IV Personality. Washington, DC, American Psychiatric Press, 1997
Pilkonis PA, Heape CL, Ruddy J, et al: Validity in the diagnosis of personality disorders: the use of the LEAD standard. Psychol Assess 3:4654, 1991
Reich W: Character Analysis (1933). New York, Simon and
Schuster, 1945
Schneider K: Psychopathic Personalities (1923). London,
Cassell, 1950
Shea MT, Yen S: Stability as a distinction between Axis I and
Axis II disorders. J Personal Disord 17:373386, 2003
Shea MT, Stout R, Gunderson J, et al: Short-term diagnostic
stability of schizotypal, borderline, avoidant, and obsessive-compulsive personality disorders. Am J Psychiatry
159:20362041, 2002
Shedler J, Westen D: Refining personality disorder diagnosis: integrating science and practice. Am J Psychiatry
161:13501365, 2004

15

Skodol AE: Classification, assessment, and differential diagnosis of personality disorders. Journal of Practical Psychology, Behavior and Health 3:261274, 1997
Skodol AE, Link BG, Shrout PE, et al: The revision of Axis V
in DSM-III-R: should symptoms have been included?
Am J Psychiatry 145:825829, 1988
Skodol AE, Gunderson JG, McGlashan TH, et al: Functional
impairment in patients with schizotypal, borderline,
avoidant, or obsessive-compulsive personality disorder. Am J Psychiatry 159:276283, 2002
Spitzer RL: Psychiatric diagnosis: are clinicians still necessary? Compr Psychiatry 24:399411, 1983
Tellegen A: Folk concepts and psychological concepts of personality and personality disorder (commentary). Psychological Inquiry 4:122130, 1993
Tyrer P: Are personality disorders well classified in DSMIV? in The DSM-IV Personality Disorders. Edited by
Livesley WJ. New York, Guilford, 1995, pp 2942
Watson D, Clark LA, Harkness AR: Structures of personality
and their relevance to psychopathology. J Abnorm Psychol 103:1831, 1994
Westen D: Divergences between clinical and research methods for assessing personality disorders: implications for
research and the evolution of Axis II. Am J Psychiatry
154:895903, 1997
Westen D, Arkowitz-Westen L: Limitations of Axis II in diagnosing personality pathology in clinical practice. Am
J Psychiatry 155:17671771, 1998
Westen D, Shedler J: Revising and assessing Axis II, part I:
developing a clinically and empirically valid assessment method. Am J Psychiatry 156:258272, 1999a
Westen D, Shedler J: Revising and assessing Axis II, part II:
toward an empirically based and clinically useful classification of personality disorders. Am J Psychiatry
156:273285, 1999b
Widiger TA: Definition, diagnosis, and differentiation. J Personal Disord 5:4251, 1991a
Widiger TA: Personality disorder dimensional models proposed for DSM-IV. J Personal Disord 5:386398, 1991b
Widiger TA: Categorical versus dimensional classification:
implications from and for research. J Personal Disord
6:287300, 1992
Widiger TA: The DSM-III-R categorical personality disorder
diagnoses: a critique and an alternative. Psychological
Inquiry 4:7590, 1993
Widiger TA: Four out of five aint bad (commentary). Arch
Gen Psychiatry 55:865866, 1998
Widiger TA: Personality disorder and Axis I psychopathology: the problematic boundary of Axis I and Axis II.
J Personal Disord 17:90108, 2003
Widiger TA, Sanderson CJ: Toward a dimensional model of
personality disorders, in The DSM-IV Personality Disorders. Edited by Livesley WJ. New York, Guilford, 1995,
pp 433458
Widiger TA, Shea T: Differentiation of Axis I and Axis II disorders. J Abnorm Psychol 100:399406, 1991

16

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Widiger TA, Frances AJ, Harris M, et al: Comorbidity among


Axis II disorders, in Personality Disorders: New Perspectives on Diagnostic Validity. Edited by Oldham JM.
Washington, DC, American Psychiatric Press, 1991
Wiggins J: Circumplex models of interpersonal behavior in
clinical psychology, in Handbook of Research Methods
in Clinical Psychology. Edited by Kendall P, Butcher J.
New York, Wiley, 1982

World Health Organization: International Classification of


Diseases, 8th Revision. Geneva, Switzerland, World
Health Organization, 1968
Zillborg G: Ambulatory schizophrenia. Psychiatry 4:149
155, 1941

2
Theories of Personality and
Personality Disorders
Amy Heim, Ph.D.
Drew Westen, Ph.D.

Personality refers to enduring patterns of cognition,


emotion, motivation, and behavior that are activated in
particular circumstances (see Mischel and Shoda 1995;
Westen 1995). This minimalist definition (i.e., one that
most personality psychologists would accept, despite
widely differing theories) underscores two important
aspects of personality. First, personality is dynamic,
characterized by an ongoing interaction of mental, behavioral, and environmental events). Second, inherent
in personality is the potential for variation and flexibility of responding (activation of specific processes under
particular circumstances). Enduring ways of responding need not be broadly generalized to be considered
aspects of personality (or to lead to dysfunction), because many aspects of personality are triggered by specific situations, thoughts, or feelings. For example, a
tendency to bristle and respond with opposition, anger,
or passive resistance to perceived demands of male authority figures may or may not occur with female authorities, peers, lovers, or subordinates. Nevertheless,

this response tendency represents an enduring way of


thinking, attending to information, feeling, and responding that is clearly an aspect of personality (and
one that can substantially affect adaptation).
Among the dozens of approaches to personality
advanced over the past century, two are of the most
widespread use in clinical practice: the psychodynamic and the cognitive-social or cognitive-behavioral. Two other approaches have gained increased interest among personality disorder researchers: trait
psychology, one of the oldest and most enduring empirical approaches to the study of normal personality;
and biological approaches, which reflect a long-standing tradition in descriptive psychiatry as well as more
recent developments in behavior genetics and neuroscience. Although most theories have traditionally
fallen into a single camp, several other approaches
are best viewed as integrative. These include Benjamins (1996a, 1996b) interpersonal approach, which
integrates interpersonal, psychodynamic, and social

Preparation of this manuscript was supported in part by NIMH MH62377 and MH62378 to the second author.

17

18

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

learning theories; Millons (1990) evolutionarysocial


learning approach, which has assimilated broadly
from multiple traditions (e.g., psychoanalytic object
relations theory); and Westens (1995, 1998) functional
domains model, which draws on psychodynamic,
evolutionary, behavioral, cognitive, and developmental research. In this chapter we briefly consider how
each approach conceptualizes personality disorders.

PSYCHODYNAMIC THEORIES
Psychoanalytic theorists were the first to generate a
concept of personality disorder (also called character
disorder, reflecting the idea that personality disorders
involve character problems not isolated to a specific
symptom or set of independent symptoms). Personality disorders began to draw considerable theoretical
attention in psychoanalysis by the middle of the twentieth century (e.g., Fairbairn 1952; Reich 1933/1978), in
part because they were common and difficult to treat,
and in part because they defied understanding using
the psychoanalytic models prevalent at the time. For
years, analysts had understood psychological problems in terms of conflict and defense using Freuds topographic model (conscious, preconscious, unconscious) or his structural model (id, ego, superego). In
classical psychoanalytic terms, most symptoms reflect
maladaptive compromises, forged outside of awareness, among conflicting wishes, fears, and moral standards. For example, a patient with anorexia nervosa
who is uncomfortable with her impulses and who
fears losing control over them may begin to starve herself as a way of demonstrating that she can control
even the most persistent of desires, hunger. Some of
the personality disorders currently identified in DSMIV (American Psychiatric Association 1994) and its update, DSM-IV-TR (American Psychiatric Association
2000), have their roots in early psychoanalytic theorizing about conflictnotably dependent, obsessivecompulsive, and to some extent histrionic personality
disorders (presumed to reflect fixations at the oral,
anal, and phallic stages, respectively).
Although some psychoanalysts have argued that a
conflict model can account for severe personality pathology (e.g., Abend et al. 1983), most analytic theorists have turned to ego psychology, object relations
theory, self psychology, and relational theories to help
understand patients with personality disorders. According to these approaches, the problems seen in patients with character disorders run deeper than maladaptive compromises among conflicting motives,

and reflect derailments in personality development reflecting temperament, early attachment experiences,
and their interaction (e.g., Balint 1969; Kernberg
1975b). Many of the DSM-IV personality disorders
have roots in these later approaches, notably schizoid,
borderline, and narcissistic personality disorders.
Psychoanalytic ego psychology focuses on the
psychological functions (in contemporary cognitive
terms, the skills, procedures, and processes involved
in self-regulation) that must be in place for people to
behave adaptively, attain their goals, and meet external demands (see Bellak et al. 1973; Blanck and Blanck
1974; Redl and Wineman 1951). From this perspective,
patients with personality disorders may have various
deficits in functioning, such as poor impulse control,
difficulty regulating affects, and deficits in the capacity for self-reflection. These deficits may render them
incapable of behaving consistently in their own best
interest or of taking the interests of others appropriately into account (e.g., they lash out aggressively
without forethought or cut themselves when they become upset).
Object relations, relational, and self psychological
theories focus on the cognitive, affective, and motivational processes presumed to underlie functioning in
close relationships (Aron 1996; Greenberg and Mitchell 1983; Mitchell 1988; Westen 1991b). From this point
of view, personality disorders reflect a number of processes. Internalization of attitudes of hostile, abusive,
critical, inconsistent, or neglectful parents may leave
patients with personality disorder vulnerable to fears
of abandonment, self-hatred, a tendency to treat themselves as their parents treated them, and so forth
(Benjamin 1996a, 1996b; Masterson 1976; McWilliams
1998). Patients with personality disorder often fail to
develop mature, constant, multifaceted representations of the self and others. As a result, they may be
vulnerable to emotional swings when significant others are momentarily disappointing, and they may
have difficulty understanding or imagining what
might be in the minds of the people with whom they
interact (Fonagy and Target 1997; Fonagy et al. 1991,
2003). Those with personality disorder often appear to
have difficulty forming a realistic, balanced view of
themselves that can weather momentary failures or
criticisms and may have a corresponding inability to
activate procedures (hypothesized to be based on loving, soothing experiences with early caregivers) that
would be useful for self-soothing in the face of loss,
failure, or threats to safety or self-esteem (e.g., Adler
and Buie 1979). A substantial body of research supports many of these propositions, particularly vis--

Theories of Personality and Personality Disorders

vis borderline personality disorder (BPD), the most


extensively studied personality disorder (e.g., Baker et
al. 1992; Gunderson 2001; Westen 1990a, 1991a).
From a psychodynamic point of view, perhaps the
most important features of personality disorders are
the following: a) they represent constellations of psychological processes, not distinct symptoms that can be
understood in isolation; b) they can be located on a continuum of personality pathology from relative health to
relative sickness; c) they can be characterized in terms
of character style, which is orthogonal to level of disturbance (e.g., a patient can have an obsessional style
but be relatively sick or relatively healthy); d) they involve both implicit and explicit personality processes,
only some of which are available to introspection (and
thus amenable to self-report); and e) they reflect processes that are deeply entrenched, often serve multiple
functions, and/or have become associated with regulation of affects and are hence resistant to change.
The most comprehensive theory that embodies
these principles is the theory of personality structure or
organization developed by Otto Kernberg (1975a, 1984,
1996). In his theory, Kernberg proposed a continuum of
pathology, from chronically psychotic levels of functioning, through borderline functioning (severe personality disorders), through neurotic to normal functioning. In Kernbergs view, people with severe personality
pathology are distinguished from people whose personality is organized at a psychotic level by their relatively intact capacity for reality testing (the absence of
hallucinations or psychotic delusions) and their relative
ability to distinguish between their own thoughts and
feelings and those of others (the absence of beliefs that
their thoughts are being broadcast on the radio; their
recognition, although sometimes less than complete,
that the persecutory thoughts in their heads are voices
from the past rather than true hallucinations, etc.).
What distinguishes individuals with severe personality
pathology from people with neurotic (that is, healthier) character structures includes 1) their more maladaptive modes of regulating their emotions through
immature, reality-distorting defenses such as denial
and projection (e.g., refusing to recognize the part they
play in generating some of the hostility they engender
from others); and 2) their difficulty in forming mature,
multifaceted representations of themselves and significant others (e.g., believing that a person they once loved
is really all bad, with no redeeming features, and is motivated only by the desire to hurt them). Kernberg refers
to these two aspects of borderline personality organization as primitive defenses and identity diffusion.
This level of severe personality disturbance, which

19

Kernberg calls borderline personality organization,


shares some features with the DSM-IVs BPD diagnosis. However, borderline personality organization is a
broader construct, encompassing patients with paranoid, schizoid, schizotypal, and antisocial personality
disorders as well as some patients who would receive a
DSM-IV diagnosis of narcissistic, histrionic, or dependent personality disorder. (Some schizotypal and borderline patients may at times fall south of the border
into the psychotic range.) Recent research supports the
notion that patients fall on a continuum of severity of
personality pathology (see Millon and Davis 1995;
Tyrer and Johnson 1996), with disorders such as paranoid and borderline personality disorder representing
more severe forms, and disorders such as obsessivecompulsive personality disorder less severe (Westen
and Shedler 1999a).
Although many of Kernbergs major contributions
have been in the understanding of borderline phenomena, his theory of narcissistic disturbance contributed
substantially to the development of the diagnosis of
narcissistic personality disorder in DSM-III (American
Psychiatric Association 1980), just as his understanding of borderline phenomena contributed to the borderline diagnosis. According to Kernberg, whereas
borderline patients lack an integrated identity, narcissistic patients are typically developmentally more advanced, in that they have been able to develop a coherent (if distorted) view of themselves. Narcissistic
phenomena, in Kernbergs view, lie on a continuum
from normal (characterized by adequate self-esteem
regulation) to pathological (narcissistic personality disorder) (Kernberg 1984, 1998). Individuals with narcissistic personality disorder need to construct a grossly
inflated view of themselves to maintain self-esteem
and may appear grandiose, sensitive to the slightest attacks on their self-esteem (and hence vulnerable to
rage or depression), or both. Not only are the conscious
self-representations of narcissistic patients inflated but
so too are the representations that constitute their ideal
selves. Actual and ideal self-representations stand in
dynamic relation to one another. Thus, one reason narcissistic patients must maintain an idealized view of
self is that they have a correspondingly grandiose view
of who they should be, a divergence that leads to tremendous feelings of shame, failure, and humiliation.
The concept of a grandiose self is central to the self
psychology of Heinz Kohut, a major theorist of narcissistic personality pathology whose ideas, like those of
Kernberg, contributed to the DSM-III diagnosis of narcissistic personality disorder (Goldstein 1985). Kohuts
theory grew out of his own and others clinical experi-

20

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

ences with patients whose problems (such as feelings of


emptiness or unstable self-esteem) did not respond well
to existing (psychoanalytic) models. Narcissistic pathology, according to Kohut, results from faulty self-development. Kohuts concept of the self refers to the nucleus
of a persons central ambitions and ideals and the talents and skills used to actualize them (Kohut 1971, 1977;
Wolf 1988). It develops through two pathways (poles)
that provide the basis for self-esteem. The first is the
grandiose selfan idealized representation of self that
emerges in children through empathic mirroring by
their parents (Mommy, watch!) and provides the nucleus for later ambitions and strivings. The second is the
idealized parent imagoan idealized representation of the
parents that provides the foundation for ideals and standards for the self. Parental mirroring allows the child to
see his reflection in the eyes of a loving and admiring
parent; idealizing a parent or parents allows the child to
identify with and become like them. In the absence of
adequate experiences with parents who can mirror the
child or serve as appropriate targets of idealization (for
example, when the parents are self-involved or abusive), the childs self-structure cannot develop, preventing the achievement of cohesion, vigor, and normal selfesteem (described by Kohut as healthy narcissism).
As a result, the child develops a disorder of the self, of
which pathological narcissism is a prototypic example.

COGNITIVE-SOCIAL THEORIES
Cognitive-social theories (Bandura 1986; Mischel
1973, 1979) offered the first comprehensive alternative
to psychodynamic approaches to personality. First developed in the 1960s, these approaches are sometimes
called social learning theory, cognitive-social learning
theory, social cognitive theory, and cognitive-behavioral theories. Cognitive-social theories developed
from behaviorist and cognitive roots. From a behaviorist perspective, personality consists of learned behaviors and emotional reactions that tend to be relatively specific (rather than highly generalized) and
tied to particular environmental contingencies. Cognitive-social theories share the behaviorist belief that
learning is the basis of personality and that personality dispositions tend to be relatively specific and
shaped by their consequences. They share the cognitive view that the way people encode, transform, and
retrieve information, particularly about themselves
and others, is central to personality. From a cognitivesocial perspective, personality reflects a constant interplay between environmental demands and the way

the individual processes information about the self


and the world (Bandura 1986).
Cognitive-social theorists have only recently begun to write about personality disorders (e.g., Beck et
al. 2003; Linehan 1993a; Pretzer and Beck 1996; Young
1990). In large part this late entrance into the study of
personality disorders reflects the assumption, initially
inherited from behaviorism, that personality is composed of relatively discrete, learned processes that are
more malleable and situation specific than implied by
the concept of personality disorder. Cognitive-social
theories focus on a number of variables presumed to
be most important in understanding personality disorders, including schemas, expectancies, goals, skills
and competencies, and self-regulation (Bandura 1986,
1999; Cantor and Kihlstrom 1987; Mischel 1973, 1979;
Mischel and Shoda 1995). Although particular theorists have tended to emphasize one or two of these
variables in explaining personality disorders, such as
the schemas involved in encoding and processing information about the self and others (Beck et al. 2003) or
the deficits in affect regulation seen in borderline patients (Linehan 1993a), a comprehensive cognitivesocial account of personality disorders would likely
address all of them.
For example, patients with personality disorders
have dysfunctional schemas that lead them to misinterpret information (as when patients with BPD misread
and misattribute peoples intentions); attend to and encode information in biased ways (as when patients
with paranoid personality disorder maintain vigilance
for perceived slights or attacks); or view themselves as
bad or incompetent (pathological self-schemas). Related to these schemas are problematic expectancies,
such as pessimistic expectations about the world, beliefs about the malevolence of others, and fears of being
mocked. Patients with personality disorders may have
pathological self-efficacy expectancies, such as the dependent patients belief that he cannot survive on his
own; the avoidant patients belief that she is likely to
fail in social circumstances, or the narcissistic patients
grandiose expectations about what he can accomplish.
Equally important are competenciesthat is, skills and
abilities used for solving problems. In social-cognitive
terms, social intelligence includes a variety of competencies that help people navigate interpersonal waters
(Cantor and Harlow 1994; Cantor and Kihlstrom 1987),
and patients with personality disorders tend to be notoriously poor interpersonal problem solvers.
Of particular relevance to severe personality disorders is self-regulation, which refers to the process of
setting goals and subgoals, evaluating ones perfor-

Theories of Personality and Personality Disorders

mance in meeting these goals, and adjusting ones behavior to achieve these goals in the context of ongoing
feedback (Bandura 1986; Mischel 1990). Problems in
self-regulation, including a deficit in specific skills,
form a central aspect of Linehans (1993a, 1993b) work
on BPD. Linehan regards emotion dysregulation as
the essential feature of BPD. The key characteristics of
emotion dysregulation include difficulty 1) inhibiting
inappropriate behavior related to intense affect, 2) organizing oneself to meet behavioral goals, 3) regulating physiological arousal associated with intense emotional arousal, and 4) refocusing attention when
emotionally stimulated (Linehan 1993b). Many of the
behavioral manifestations of BPD (e.g., cutting) can be
viewed as consequences of emotional dysregulation.
Deficits in emotion regulation lead to other problems,
such as difficulties with interpersonal functioning and
with the development of a stable sense of self.
According to another cognitive-behavioral approach, Becks cognitive theory (Beck 1999; Beck et al.
2003; Pretzer and Beck 1996), dysfunctional beliefs
constitute the primary pathology involved in the personality disorders (Beck et al. 2001), which are viewed
as pervasive, self-perpetuating cognitive-interpersonal cycles (Pretzer and Beck 1996, p. 55). Becks theory highlights three aspects of cognition: 1) automatic
thoughts (beliefs and assumptions about the world,
the self, and others); 2) interpersonal strategies; and
3) cognitive distortions (systematic errors in rational
thinking). Beck and colleagues have described a
unique cognitive profile characteristic of each of the
DSM-IV personality disorders. For example, an individual diagnosed with schizoid personality disorder
would have a view of himself as a self-sufficient loner,
a view of others as unrewarding and intrusive, and a
view of relationships as messy and undesirable, and
his primary interpersonal strategy would involve
keeping his distance from other people (Pretzer and
Beck 1996). He would use cognitive distortions that
minimize his recognition of how relationships with
others can be sources of pleasure. A recent study of
dysfunctional beliefs (as assessed by the Personal Beliefs Questionnaire [A.T. Beck, J.S. Beck, unpublished
assessment instrument, The Beck Institute for Cognitive Therapy and Research, Bala Cynwyd, Pennsylvania, 1991]) provides some initial support for the link
between particular beliefs and the DSM-IV personality disorders (Beck et al. 2001).
Building on Becks cognitive theory, Young and colleagues (Young and Gluhoski 1996; Young and Lindemann 2002; Young et al. 2003) have added a fourth
level of cognition: early maladaptive schemas, which

21

they have defined as broad and pervasive themes regarding oneself and ones relationships with others,
developed during childhood and elaborated throughout ones life (Young and Lindemann 2002, p. 95). The
authors distinguish these schemas from automatic
thoughts and underlying assumptions, noting that the
schemas are associated with greater levels of affect, are
more pervasive, and involve a strong interpersonal aspect. Young and colleagues have identified 16 early
maladaptive schemas, each of which comprises cognitive, affective, and behavioral components. They have
also identified three cognitive processes involving
schemas that define key features of personality disorders: schema maintenance, which refers to the processes
by which maladaptive schemas are rigidly upheld
(e.g., cognitive distortions, self-defeating behaviors);
schema avoidance, which refers to the cognitive, affective, and behavioral ways individuals avoid the negative affect associated with the schema; and schema compensation, which refers to ways of overcompensating
for the schema (e.g., becoming a workaholic in response to a schema of self as failure).
Mischel and Shoda (1995) have offered a compelling social-cognitive account of personality that focuses on ifthen contingenciesthat is, conditions
that activate particular thoughts, feelings, and behaviors. Although they have not linked this model to personality disorders, one could view personality disorders as involving a host of rigid, maladaptive ifthen
contingencies. For example, for some patients, the first
hints of trouble in a relationship may activate concerns
about abandonment. These in turn may elicit anxiety
or rage, to which the patient responds with desperate
attempts to lure the person back that often backfire
(such as manipulative statements and suicidal gestures). From an integrative psychodynamic-cognitive
viewpoint, Horowitz (1988, 1998) offered a model that
similarly focused on the conditions under which certain states of mind become active, which he has tied
more directly to a model of personality disorders; and
Wachtel (1977, 1997) has similarly described cyclical
psychodynamics, in which people manage to elicit
from others precisely the kind of reactions of which
they are the most vigilant and afraid.

TRAIT THEORIES
Trait psychology focuses less on personality processes
or functions than do psychodynamic or cognitivesocial approaches, and hence has not generated an approach to treatment, although it has generated highly

22

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

productive empirical research programs. Traits are


emotional, cognitive, and behavioral tendencies on
which individuals vary (e.g., the tendency to experience negative emotions). According to Gordon Allport (1937), who pioneered the trait approach to personality, the concept of trait has two separate but
complementary meanings: it is both an observed tendency to behave in a particular way and an inferred
underlying personality disposition that generates this
behavioral tendency. In the empirical literature, traits
have largely been defined operationally, as the average of a set of self-report items designed to assess a
given trait (e.g., items indicating a tendency to feel
anxious, sad, ashamed, guilty, self-doubting, and angry that all share a common core of negative affectivity or neuroticism).
Researchers have recently begun recasting personality disorders in terms of the most prominent contemporary trait theory, the Five-Factor Model of personality (FFM; McCrae and Costa 1997; Widiger 2000;
Widiger and Costa 1994). (We address other trait models that have been more closely associated with biological theories later.) The FFM is a description of the way
personality descriptors tend to covary and hence can
be understood in terms of latent factors (traits) identified via factor analysis. Based on the lexical hypothesis
of personalitythat important personality attributes
will naturally find expression in words used in everyday languagethe FFM emerged from factor analysis
of adjectival descriptions of personality originally selected from Websters Unabridged Dictionary (Allport
and Odbert 1936). Numerous studies, including crosscultural investigations, have found that when participants in nonclinical (normal) samples are asked to rate
themselves on dozens or hundreds of adjectives or
brief sentences, the pattern of self-descriptions can often be reduced to five overarching constructs (Costa
and McCrae 1997; Goldberg 1993): 1) neuroticism or
negative affect (how much they tend to be distressed);
2) extraversion or positive affect (the extent to which
they tend to be gregarious, high-energy, and happy);
3) conscientiousness; 4) agreeableness; and 5) openness to experience (the extent to which they are open
to emotional, aesthetic, and intellectual experiences).
McCrae and Costa (1990, 1997) proposed a set of
lower-order traits, or facets, within each of these
broadband traits that can allow a more discriminating
portrait of personality. Thus, an individuals personality profile is represented by a score on each of the five
factors plus scores on six lower-order facets or subfactors within each of these broader constructs (e.g., anxiety and depression as facets of neuroticism). Advo-

cates of the FFM argue that personality disorders


reflect extreme versions of normal personality traits,
so that the same system can be used for diagnosing
normal and pathological personality. From the perspective of the FFM, personality disorders are not discrete entities separate and distinct from normal personality. Rather, they represent extreme variants of
normal personality traits or blends thereof.
In principle, one could classify personality disorders in one of two ways using the FFM. The first, and
that more consistent with the theoretical and psychometric tradition within which the FFM developed, is
simply to identify personality pathology by extreme
values on each of the five factors (and perhaps on their
facets). For example, extremely high scores on the neuroticism factor and its facets (anxiety, hostility, depression, self-consciousness, impulsivity, and vulnerability) all represent aspects of personality pathology.
Whether this strategy is appropriate for all factors and
facets, and when to consider extreme responses on one
or both poles of a dimension pathological, are matters
of debate. Extreme extraversion, for example, may or
may not be pathological, depending on the social milieu and the persons other traits. Similarly, extreme
openness to experience could imply a genuinely open
attitude toward emotions, art, and so forth or an uncritical, flaky, or schizotypal cognitive style. The advantages of this approach, however, are that it integrates the understanding and assessment of normal
and pathological personality and that it establishes dimensions of personality pathology using well-understood empirical procedures (factor analysis).
Another way to proceed using the FFM is to translate clinically derived categories into five-factor language (Coker et al. 2002; Lynam and Widiger 2001; Widiger and Costa 1994). For example, Widiger and
colleagues (2002) described antisocial personality disorder (ASPD) as combining low agreeableness with
low conscientiousness. Because analysis at the level of
five factors often lacks the specificity to characterize
complex disorders such as BPD (high neuroticism plus
high extraversion), proponents of the FFM have often
moved to the facet level. Thus, whereas all six neuroticism facets (anxiety, hostility, depression, self-consciousness, impulsivity, and vulnerability) are characteristic of patients with BPD, patients with avoidant
personality disorder are characterized by only four of
these facets (anxiety, depression, self-consciousness,
and vulnerability). Similarly, Widiger and colleagues
(1994, 2002) described obsessive-compulsive personality disorder as primarily an extreme, maladaptive
variant of conscientiousness. They add, however, that

Theories of Personality and Personality Disorders

obsessive-compulsive patients tend to be low on the


compliance and altruism facets of agreeableness (i.e.,
they are oppositional and stingy) and low on some of
the facets of openness to experience as reflected in being closed to feelings and closed to values (i.e., morally inflexible). Numerous studies have shown predicted links between DSM-IV Axis II disorders and
FFM factors and facets (Axelrod et al. 1997; Ross et al.
2002; Trull et al. 2001), although other studies have
found substantial overlap among the FFM profiles of
patients with very different disorders (e.g., borderline
and obsessive-compulsive) using major FFM selfreport inventories (Morey et al. 2002).

BIOLOGICAL PERSPECTIVES
The first biological perspectives on personality disorders, which influenced the current Axis II classification,
stemmed from the observations of the pioneering psychiatric taxonomists in the early twentieth century, notably Bleuler (1911/1950) and Kraepelin (1896/1919).
These authors and others noticed, for example, that the
relatives of schizophrenic patients sometimes appeared
to have attenuated symptoms of the disorder that endured as personality traits, such as interpersonal and
cognitive peculiarity. More recently, researchers have
used the methods of trait psychology (particularly the
reliance on self-report questionnaires and factor analysis) to study personality disorders from a biological
viewpoint. In some cases, they have developed item
sets with biological variables in mind (e.g., neurotransmitters and their functions) or have reconsidered patterns of covariation among different traits in light of hypothesized neurobiological systems or circuits. In other
cases, they have applied behavior-genetic approaches
to study personality traits (as well as DSM-IV disorders). We explore each of these approaches in turn. (Researchers are just beginning to use neuroimaging to
study personality disorders, particularly BPD [e.g.,
Herpertz et al. 2001], but the results at this point are preliminary, and hence we do not address them further
here.)

Traits and Neural Systems


Siever and Davis (1991) provided one of the first attempts to reconsider the personality disorders from a
neurobiology perspective. They proposed a model
based on core characteristics of Axis I disorders relevant to personality disorders and related these characteristics to emerging knowledge of their underlying

23

neurobiology. They focused on cognitive/perceptual


organization (schizophrenia and other psychotic disorders); impulsivity/aggression (impulse control disorders); affective instability (mood disorders); and
anxiety/inhibition (anxiety disorders). Conceptualized in dimensional terms, Axis I disorders such as
schizophrenia represent the extreme end of a continuum. Milder abnormalities can be seen in patients
with personality disorder, either directly (as subthreshold variants) or through their influence on
adaptive strategies (coping and defense).
Siever and Davis linked each dimension to biological correlates and indicators, some presumed to be
causal and others to provide markers of underlying biological dysfunction (e.g., eye movement dysfunction
in schizophrenia, which is also seen in individuals
with schizotypal personality disorder and in nonpsychotic relatives of schizophrenic probands). They also
pointed to suggestive data on neurotransmitter functioning that might link Axis II disorders with Axis I
syndromes such as depression. More recently, Siever
and colleagues (New and Siever 2002; Siever et al.
2003) proposed an approach to BPD that tries to circumvent the problems created by the heterogeneity of
the diagnosis by examining the neurobiology of specific dimensions thought to underlie the disorder (endophenotypes), especially impulsive aggression and
affective instability.
The major attempt thus far to develop a trait model
of personality disorders based on a neurobiological
model is Cloningers seven-factor model of personality (Cloninger 1998; Cloninger et al. 1993). In his
model, Cloninger divided personality structure into
two domains: temperament (automatic associative responses to basic emotional stimuli that determine habits and skills) and character (self-aware concepts that
influence voluntary intentions and attitudes) (Cloninger 1998, p. 64). According to Cloninger, each of
these domains is defined by a mode of learning and
the underlying neural systems involved in that learning: temperament is associated with associative/procedural learning, and character is associated with insight learning. The temperament domain includes
four dimensions, each theoretically linked to particular neurotransmitter systems: 1) novelty seeking (exploration, extravagance, impulsivity), associated with
dopamine; 2) harm avoidance (characterized by pessimism, fear, timidity), associated with serotonin and
GABA (-aminobutyric acid); 3) reward dependence
(sentimentality, social attachment, openness), associated with norepinephrine and serotonin; and 4) persistence (industriousness, determination, ambitiousness,

24

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

perfectionism), associated with glutamate and serotonin (Cloninger 1998, p. 70). The character domain includes three dimensions: 1) self-directedness (responsibility, purposefulness, self-acceptance), considered
the major determinant of the presence or absence of
personality disorder (Cloninger et al. 1993, p. 979); 2)
cooperativeness (empathy, compassion, helpfulness);
and 3) self-transcendence (spirituality, idealism, enlightenment).
Cloninger (1998) proposed that all personality disorders are low on the character dimensions of selfdirectedness and cooperativeness. What distinguishes
patients with different disorders are their more specific profiles. In broad strokes, the Cluster A personality disorders (schizotypal, schizoid, paranoid) are associated with low reward dependence; the Cluster B
personality disorders (borderline, antisocial, narcissistic, histrionic) are associated with high novelty seeking; and the Cluster C personality disorders (dependent, avoidant, obsessive-compulsive) are associated
with high harm avoidance. Individual personality disorders may be described more fully by profiles obtained from Cloningers self-report Temperament and
Character Inventory (Cloninger and Svrakic 1994). For
example, BPD would consist of high harm avoidance,
high novelty seeking, and low reward dependence as
well as low scores on the character dimensions.
More recently, a dimensional neurobehavioral
model was offered by Depue, Lenzenweger, and colleagues (e.g., Depue and Collins 1999; Depue and
Lenzenweger 2001). Their model regards personality
disorders as emergent phenotypes arising from the interaction of basic neurobehavioral systems that underlie major personality traits (Depue and Lenzenweger
2001, p. 165). Through an extensive examination of the
psychometric literature on the structure of personality
traits as well as a theoretical analysis of the neurobehavioral systems likely to be relevant to personality
and personality dysfunction, they identified five trait
dimensions that may account for the range of personality disorder phenotypes. They labeled these five
traits 1) agentic extraversion (reflecting both the activity and gregariousness components of extraversion);
2) neuroticism; 3) affiliation; 4) nonaffective constraint
(the opposite pole of which is impulsivity); and 5) fear.
For example, the neurobehavioral system underlying
the trait of agentic extraversion is positive incentive
motivation, which is common to all mammalian species and involves positive affect and approach motivation. The dopaminergic system has been strongly implicated in incentive-motivated behavior, such that
individual differences in the former predict differ-

ences in the latter. Research on this model is just beginning, but the model is promising in its integration of
research on neural systems involved in fundamental
functions common to many animal species (such as
approach, avoidance, affiliation with conspecifics, and
inhibition of punished behavior) with individual differences research in personality psychology.

Behavior-Genetic Approaches
The vast majority of behavior-genetic studies of personality have focused on normal personality traits,
such as those that compose the FFM and Eysencks
(1967, 1981) three-factor model (extraversion, neuroticism, and psychoticism). These studies have generally
shown moderate to high heritability (30%60%) for a
range of personality traits (Livesley et al. 1993; Plomin
and Caspi 1999) relevant to personality disorders. The
most frequently studied traits, extraversion and neuroticism, have produced heritability estimates of 54%
74% and 42%64%, respectively (Eysenck 1990).
Behavior-genetic data are proving increasingly
useful in both etiological and taxonomic work (e.g.,
Krueger 1999; Livesley et al. 1998). Livesley and colleagues (2003) noted that behavior-genetic data can
help address the persistent lack of consensus among
trait psychologists regarding which traits to study by
helping them study the causes of trait covariation (as
opposed to simply describing it). Establishing congruence between a proposed phenotypic model of
personality traits and the genetic structure underlying
it would support the validity of a proposed factor
model. The same holds true for models of personality
disorders. To test this approach, Livesley et al. (1998)
administered the Dimensional Assessment of Personality PathologyBasic Questionnaire to a large sample of individuals with and without personality disorders, including twin pairs. This self-report measure
consists of 18 traits considered to underlie personality
disorder diagnoses (e.g., identity problems, oppositionality, social avoidance). Factor analysis indicated a
four-factor solution: emotional dysregulation, dissocial behavior, inhibition, and compulsivity. Results
showed high congruence for all four factors between
the phenotypic and behavior-genetic analyses, indicating strong support for the proposed factor solution.
In addition, the data showed substantial residual heritability for many lower-order traits, suggesting that
these traits likely are not simply components of the
higher-order factors but include unique components
(specific factors) as well. Krueger and colleagues (e.g.,
Krueger 1999) have similarly found, using structural

Theories of Personality and Personality Disorders

equation modeling with a large twin sample, that


broadband internalizing and externalizing personality factors account for much of the variance in many
common Axis I disorders (e.g., mood, anxiety, and
substance use) and that genetic and environmental
sources of variance are associated with many of both
the higher- and lower-order factors they identified.
Compared with research on normal personality
traits (as well as many Axis I disorders), behaviorgenetic studies of personality disorders are relatively
rare. The most common designs have been family
studies in which researchers begin with the personality disorder proband and then assess other family
members. The major limitation of this method is that
familial aggregation of disorders can support either
genetic or environmental causes. As in all behaviorgenetic research, twin and adoption studies provide
more definitive data. Most of these studies have examined only a subset of the DSM personality disorders, particularly schizotypal, antisocial, and borderline personality disorders. These disorders appear to
reflect a continuum of heritability, with schizotypal
most strongly linked to genetic influences, antisocial
linked both to environmental and genetic variables,
and borderline showing the smallest estimates of heritability in the majority of studies (see Nigg and Goldsmith 1994).
Research on the heritability of schizotypal personality disorder provides the clearest evidence of a genetic component to a personality disorder. (Schizotypal personality disorder is defined by criteria such
as odd beliefs or magical thinking, unusual perceptual
experiences, odd thinking and speech, suspiciousness,
inappropriate or constricted affect, and behavior or
appearance that is odd or eccentric.) As mentioned
earlier, Bleuler and Kraepelin noted peculiarities in
language and behavior among some relatives of their
schizophrenic patients. Bleuler called this presentation latent schizophrenia and considered it to be a
less severe and more widespread form of schizophrenia. Further research into the constellation of symptoms characteristic of relatives of schizophrenic patients ultimately resulted in the creation of the DSM
diagnosis of schizotypal personality disorder (Spitzer
et al. 1979). A genetic relationship between schizophrenia and schizotypal personality disorder is now
well established (Kendler and Walsh 1995; Lenzenweger 1998). In one study, Torgersen (1984) found that
33% (7 of 21) of identical co-twins had schizotypal personality disorder, whereas only 4% (1 of 23) of fraternal co-twins shared the diagnosis. Data from a later
twin study (Torgersen et al. 2000), which used struc-

25

tural equation modeling, estimated heritability at 0.61.


ASPD, in contrast, appears to have both genetic
and environmental roots, as documented in adoption
studies (Cadoret et al. 1995). An adult adoptee whose
biological parent has an arrest record for antisocial behavior is four times more likely to have problems with
aggressive behavior than a person without a biological
vulnerability. At the same time, a person whose adoptive parent has ASPD is more than three times more
likely to develop the disorder, regardless of biological
history. As is the case with other behavior-genetic
findings, twin studies suggest that environmental genetic factors grow more predictive as individuals get
older (Lyons et al. 1995). In considering the data on
ASPD and other personality disorders, however, it is
important to remember that all estimates of heritability are sample dependent. Turkheimer et al. (2003) recently found, for example, that genes account for most
of the variability in IQ among middle-class children
but that over 60% of the variance in IQ in samples
from low socioeconomic backgrounds reflects shared
environment. Socioeconomic status may similarly
moderate the relation between genes and environment and antisocial behavior.
Data on the behavioral genetics of BPD are mixed.
Several studies have found only modest evidence of
heritability (e.g., Dahl 1993; Nigg and Goldsmith
1994; Reich 1989). A rare twin study conducted by
Torgersen (1984) failed to find evidence for the genetic transmission of the disorder, although the sample was relatively small. A more recent twin study by
Torgersen et al. (2000) focused on the heritability of
several personality disorders, finding a substantial
genetic component to several personality disorders,
with most heritability estimates between 0.50 and
0.60, including BPD. Increasingly, researchers are suggesting that specific components of BPD may have
higher heritability than the BPD diagnosis taken as a
whole. For example, several authors (Nigg and Goldsmith 1994; Widiger and Frances 1994) suggest that
neuroticism, which is highly heritable, is at the core of
many borderline features (e.g., negative affect and
stress sensitivity). Other components of BPD have
shown substantial heritability as well (e.g., problems
with identity, impulsivity, affective lability) (Livesley
et al. 1993; Skodol et al. 2002).
A caveat worth mentioning, however, is that behavior-genetic studies that systematically measure environmental influences directly (e.g., measuring developmental toxins such as sexual abuse), rather than
deriving estimates of shared and nonshared environment statistically from residual terms, often obtain

26

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

very different estimates of environmental effects, and


this may well be the case with many personality disorders. For example, if one child in a family responds to
sexual abuse by becoming avoidant and constricted
and another responds to the same experience by becoming borderline and impulsive, researchers will
mistakenly concludeunless they actually measured
developmental variablesthat shared environment
has no effect, because a shared environmental event
led to nonshared responses (see Turkheimer and Waldron 2000; Westen 1998). Recent work by Caspi, Moffitt, and colleagues (2002) showing genes and environmental events (e.g., sexual abuse) interacting in
predicting subsequent personality and psychopathology emphasize the same point.

INTEGRATIVE THEORIES
Of all the disorders identified in DSM-IV-TR, the personality disorders are likely to be among those that
most require biopsychosocial perspectives. They are
also disorders for which we may gain substantially by
integrating data from both clinical observation and research, from classical theories of personality that delineate personality functions, and from more contemporary research that emphasizes traits. The emergence
of several integrative models is thus perhaps not surprising. We briefly describe three such models in the
following discussion: Millons evolutionarysocial
learning model, Benjamins interpersonal model, and
Westens functional-domains model.

Millons EvolutionarySocial Learning Model


Millon developed a comprehensive model of personality and personality disorders that he initially framed in
social learning terms (Millon 1969), describing personality in terms of three polarities: pleasure/pain, self/
other, and passive/active. These polarities reflect the
nature of reinforcement that controls the persons behavior (rewarding or aversive), the source or sources
that provide reinforcement (oneself or others), and the
instrumental behaviors and coping strategies used to
pursue reinforcement (active or passive). Millon (Davis
and Millon 1999; Millon 1990; Millon and Davis 1996;
Millons Chapter 14, Sociocultural Factors, this volume) eventually reconceptualized his original theory in
evolutionary terms. In doing so, he added a fourth polarity, thinking/feeling, which reflects the extent to
which people rely on abstract thinking or intuition.
Millons reconceptualized theory outlined four ba-

sic evolutionary principles consistent with the polarities described by his earlier theory: 1) aims of existence,
which refer to life enhancement and life preservation,
and which are reflected in the pleasure/pain polarity;
2) modes of adaptation, which he described in terms of
accommodation to, versus modification of, the environment (whether one adjusts or tries to adjust the world,
particularly other people) and which are reflected in the
passive/active polarity; 3) strategies of replication or
reproduction, which refer to the extent to which the
person focuses on individuation or nurturance of others and which are reflected in the self/other polarity;
and 4) processes of abstraction, which refer to the ability for symbolic thought and which are represented by
the thinking/feeling polarity.
Millon identified 14 personality prototypes that
can be understood in terms of the basic polarities. For
example, patients with schizoid personality disorder
tend to have little pleasure, to have little involvement
with others, to be relatively passive in their stance to
the world, and to rely on abstract thinking over intuition. In contrast, patients with histrionic personality
disorder are pleasure seeking, interpersonally focused
(although in a self-centered way), highly active, and
short on abstract thinking. Millons theory led to the
distinction between avoidant and schizoid personality
disorder in DSM-III. Whereas schizoid personality
disorder represents a passive-detached personality
style, avoidant personality disorder represents an active-detached style characterized by active avoidance
motivated by avoidance of anxiety. Millon also developed both a comprehensive measure to assess the
DSM personality disorders and his own theory-driven
personality disorder classification, the Millon Clinical
Multiaxial Inventory (Millon and Davis 1997). The instrument, now in its third edition, has been used in
hundreds of studies and is widely used as an assessment tool in clinical practice (e.g., Espelage et al. 2002;
Kristensen and Torgersen 2001).

Benjamins Interpersonal Model


Benjamins (1993, 1996a, 1996b) interpersonal theory,
called Structural Analysis of Social Behavior (SASB),
focuses on interpersonal processes in personality and
psychopathology and their intrapsychic causes, correlates, and sequelae. Influenced by Sullivans (1953) interpersonal theory of psychiatry, by object relations
approaches, and by research using the interpersonal
circumplex (e.g., Kiesler 1983; Leary 1957; Schaefer
1965), the SASB is a three-dimensional circumplex
model with three surfaces, each of which represents

Theories of Personality and Personality Disorders

a specific focus. The first surface focuses on actions directed at a person (e.g., abuse by a parent toward the
patient). A second surface focuses on the persons response to real or perceived actions by the other (e.g.,
recoiling from the abusive parent). The third focus is
on the persons actions toward him- or herself, or what
Benjamin calls the introject (e.g., self-abuse). The notion behind the surfaces is that the first two are interpersonal and describe the kinds of interaction patterns
(self with other) in which the patient engages with significant others (e.g., parents, attachment figures, therapists). The third surface represents internalized attitudes and actions toward the self (e.g., self-criticism
that began as criticism from parents). According to
Benjamin, children learn to respond to themselves and
others by identifying with significant others (acting
like them), recapitulating what they experienced with
significant others (e.g., eliciting from others what they
experienced before), and introjecting others (treating
themselves as others have treated them).
As with all circumplex models, each surface has
two axes that define its quadrants. In the SASB (as in
other interpersonal circumplex models), love and hate
represent the two poles of the horizontal axis. Enmeshment and differentiation are the endpoints of the
vertical axis. The SASB offers a translation of each of
the DSM Axis II criteria (and disorders) into interpersonal terms (Benjamin 1993, 1996b). In this respect, it
has two advantages. First, it reduces comorbidity
among disorders by specifying the interpersonal antecedents that elicit the patients responses. For example, maladaptive anger is characteristic of many of the
DSM-IV personality disorders but has different interpersonal triggers and meanings (Benjamin 1993).
Anger in patients with BPD often reflects perceived
neglect or abandonment. Anger in narcissistic personality disorder tends to follow from perceived slights
or failures of other people to give the patient everything he or she wants (entitlement). Anger in patients
with ASPD is often cold, detached, and aimed at controlling the other person. Second, the SASB model is
able to represent multiple, often conflicting aspects of
the way patients with a given disorder behave (or
complex, multifaceted aspects of a single interpersonal interaction) simultaneously. Thus, a single angry outburst by a borderline patient could reflect an
effort to get distance from the other, to hurt the other,
and to get the other to respond and hence be drawn
back into the relationship. Benjamin has devised several ways of operationalizing a persons dynamics or
an interpersonal interaction (e.g., in a therapy hour),
ranging from direct observation and coding of behav-

27

ior to self-report questionnaires, all of which yield descriptions using the same circumplex model.

Westens Functional-Domains Model


Westen (1995, 1996, 1998) described a model of domains of personality functioning that draws substantially on psychoanalytic clinical theory and observation as well as on empirical research in personality,
cognitive, developmental, and clinical psychology.
Although some aspects of the model are linked to research on etiology, the model is less a theory of personality disorders than an attempt to delineate and
systematize the major elements of personality that define a patients personality, whether or not the patient
has a personality disorder. The model differs from
trait approaches in its focus on personality processes
and functions (e.g., the kinds of affect regulation strategies the person uses, the ways she represents the self
and others mentally, as well as more behavioral dispositions, such as whether she engages in impulsive or
self-destructive behavior). However, it shares with
trait approaches the view that a single model should
be able to accommodate relatively healthy as well as
relatively disturbed personality styles and dynamics.
The model suggests that a systematic personality
case formulation must answer three questions, each
composed of a series of subquestions or variables that
require assessment: 1) What does the person wish for,
fear, and value, and to what extent are these motives
conscious or unconscious, collaborating or conflicting? 2) What psychological resourcesincluding cognitive processes (e.g., intelligence, memory, intactness
of thinking processes), affects, affect regulation strategies (conscious coping strategies and unconscious defenses), and behavioral skillsdoes the person have
at his or her disposal to meet internal and external demands? 3) What is the persons experience of the self
and others, and how able is the individualcognitively, emotionally, motivationally, and behaviorally
to sustain meaningful and pleasurable relationships?
From a psychodynamic perspective, these questions correspond roughly to the issues raised by classical psychoanalytic theories of motivation and conflict (Brenner 1982); ego-psychological approaches to
adaptive functioning; and object-relational, self-psychological, attachment, and contemporary relational
(Aron 1996; Mitchell 1988) approaches to understanding peoples experience of self with others. Each of
these questions and subdimensions, however, is also
associated with a number of research traditions in personality, clinical, cognitive, and developmental psy-

28

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

chology (e.g., on the development of childrens representations of self, representations of others, moral
judgment, attachment styles, ability to tell coherent
narratives) (see Damon and Hart 1988; Fonagy et al.
2002; Harter 1999; Livesley and Bromley 1973; Main
1995; Westen 1990a, 1990b, 1991b, 1994). Westen and
Shedler (1999a) used this model as a rough theoretical
guide to ensure comprehensive coverage of personality domains in developing items for the ShedlerWesten Assessment Procedure Q-Sort, a personality
pathology measure for use by expert informants, although the model and the measure are not closely
linked (i.e., one does not require the other).
From this point of view, individuals with particular personality disorders are likely to be characterized
by a) distinct constellations of motives and conflicts,
such as chronic worries about abandonment in BPD or
a conflict between the wish for and fear of connectedness to others in avoidant personality disorder; b) deficits in adaptive functioning, such as poor impulse
control, lack of self-reflective capacities (see Fonagy
and Target 1997), and difficulty regulating affect (Linehan 1993a; Westen 1991a) in BPD or subclinical cognitive disturbances in schizotypal personality disorder;
and c) problematic ways of thinking, feeling, and behaving toward themselves and significant others, such
as a tendency to form simplistic, one-dimensional representations of the self and others, to misunderstand
why people (including the self) behave as they do, and
to expect malevolence from other people (characteristics seen in patients with many personality disorders,
such as paranoid, schizoid, and borderline) (Kernberg
1975a, 1984; Westen 1991a). In this model, a persons
level of personality healthsickness (from severe personality disorder to relatively healthy functioning),
which can be assessed reliably using a personality
health prototype or a simple rating of level of personality organization derived from Kernbergs work
(Westen and Muderrisoglu 2003; Westen and Shedler
1999b), reflects his or her functioning in each of these
three domains.
People who do not have severe enough pathology
to receive a personality disorder diagnosis can similarly be described using this approach. For example, a
successful male executive presented for treatment
with troubles in his marriage and his relationships at
work, as well as low-level feelings of anxiety and depression. None of these characteristics approached criteria for a personality disorder (or any Axis I disorders, except the relatively nondescript diagnosis of
adjustment disorder with mixed anxious and depressed mood). Using this model, one would note that

he was competitive with other people, a fact of which


he was unaware (Question 1); had impressive capacities for self-regulation but was intellectualized and
afraid of feelings and often used his enjoyment of his
work as a way of retreating from his family (Question
2); and had surprisingly noncomplex representations
of others minds (for a person who could solve noninterpersonal problems in complex ways) and consequently would often became angry and attack at work
without stopping to empathize with the other persons perspective (Question 3). This description is, of
course, highly oversimplified, but it gives a sense of
how the model can be used to describe personality dynamics in patients without a diagnosable personality
disorder (Westen 1998; Westen and Shedler 1999b).

CASE EXAMPLE
To see how some of the models discussed here operate
in practice, consider the following brief case description:
Mr. A was a man in his early 20s who came to treatment for lifelong problems with depression, anxiety,
and feelings of inadequacy. He was a kind, introspective, sensitive man who nevertheless had tremendous difficulty making friends and interacting comfortably with people. He was constantly worried that
he would misspeak, he would ruminate after conversations about what he had said and the way he was
perceived, and he had only one or two friends with
whom he felt comfortable. He wanted to be closer to
people, but he was frightened that he would be rejected and was afraid of his own anger in relationships. While interacting with people (including his
therapist), he would often have a running commentary with them in his mind, typically filled with aggressive content. He was in a 2-year relationship with
a woman who was emotionally and physically very
distant, whom he saw twice a month and with whom
he rarely had sex. Prior to her, his sexual experiences
had all been anxiety provoking and short lived, in every sense.
Mr. A tended to be inhibited in many areas of his
life. He was emotionally constricted and seemed
particularly uncomfortable with pleasurable feelings. He tended to speak in intellectualized terms
about his life and history and seemed afraid of affect.
He felt stifled in his chosen profession, which did not
allow him to express many of his intellectual abilities
or creative impulses. He alternated between overcontrol of his impulses, which was his modal stance
in life, and occasional breakthroughs of poorly
thought-out, impulsive actions (as when he bought
an expensive piece of equipment with little forethought about how he would pay for it).

Theories of Personality and Personality Disorders

Mr. A came from a working class family in Boston


and had lost his father, a policeman, as a young boy. He
was reared by his mother and later by a stepfather with
whom he had a positive relationship. He also described
a good relationship with his mother, although she, like
several members of her extended family, struggled
with depression, and she apparently suffered a lengthy
major depressive episode after her husbands death.

For purposes of brevity, we briefly explicate this


case from two theoretical standpoints that provide
very different approaches to case formulation: the
FFM and the functional-domains viewpoint. (In clinical practice, a functional-domains account and a psychodynamic account are similar, because the former
reflects an attempt to systematize and integrate with
empirical research [and minimal jargon] the major domains emphasized by classical psychoanalytic, egopsychological, and object-relational/self-psychological/relational approaches.)
From a five-factor perspective, the most salient features of Mr. As personality profile were his strong elevations in neuroticism and introversion (low extraversion). He was high on most of the facets of neuroticism,
notably anxiety, depression, anger, self-consciousness,
and vulnerability. He was low on most facets of extraversion as well, particularly gregariousness, assertiveness, activity, and happiness. This combination of high
negative affectivity and low positive affectivity, which
left him vulnerable to feelings of depression, captures
his anxious, self-conscious social avoidance.
No other broadband factors describe Mr. A adequately, although specific FFM facets provide insight
into his personality. He was moderately high in agreeableness, being compliant, modest, and tender-minded;
however, he was not particularly high on trust, altruism, or straightforwardness (reflecting his tendency to
behave passive-aggressively). He was moderately conscientious, showing moderate scores on the facets of orderliness and discipline. He similarly showed moderate openness to experience, being artistically oriented
but low on comfort with feelings. His scores on facets
such as intellectual curiosity would likely be moderate,
reflecting both an interest and an inhibition. Indeed, a
tendency to receive moderate scores because of opposing dynamics would be true of his facet scores on several traits, such as achievement orientation.
A functional domains perspective would offer a
similar summary diagnosis to that of a psychodynamic approach, along with a description of his functioning on the three major domains outlined in the
model. In broadest outline, from this point of view
Mr. A had a depressive, avoidant, and obsessional

29

personality style organized at a low-functioning neurotic level. In other words, he did not have a personality disorder, as evidenced by his ability to maintain
friendships and stable employment, but he had considerable psychological impediments to love, work,
and life satisfaction, with a predominance of depressive, avoidant, and obsessional dynamics.
With respect to motives and conflicts (and interpersonal issues, around which many of his conflicts
centered), Mr. A had a number of conflicts that impinged on his capacity to lead a fulfilling life. He
wanted to connect with people, but he was inhibited
by social anxiety, feelings of inadequacy, and an
undercurrent of anger toward people that he could
not directly express (which emerged in his running
commentaries in his mind). Although he worried
that he would fail others, he always felt somehow
unfulfilled in his relationships with them and could
be subtly critical. He likely had high standards with
which he compared himself and others and against
which both frequently fell short. He also had trouble
handling his anger, aggressive impulses, and desires
for self-assertion. He would frequently behave in
passive or self-punitive ways rather than appropriately asserting his desires or expressing his anger.
This pattern contributed in turn to a lingering hostile
fantasy life and a tendency at times to behave passive-aggressively.
Sex was particularly conflictual for Mr. A, not only
because it forced him into an intimate relationship
with another person but because of his feelings of
inadequacy, his discomfort in looking directly at a
womans body (because of his associations to sex and
womens bodies), and his worries that he was homosexual. When with a woman, he frequently worried
that he would accidentally touch her anus and be repulsed, although interestingly, his sexual fantasies
(and humor) had a decidedly anal tone. Homosexual
images would also jump into his mind in the middle
of sexual activity, which led to considerable anxiety.
With respect to adaptive resources, Mr. A had a
number of strengths, notably his impressive intellect,
a dry sense of humor, a capacity to introspect, and an
ability to persevere. Nevertheless, his overregulation
of his feelings and impulses left him vulnerable to
breakthroughs of anger, anxiety, and impulsive action.
He distanced himself from emotion, in an effort both
to regulate anxiety and depression and to regulate excitement and pleasure, which seemed to him both undeserved and threatening.
With respect to his experience of self and relationships, Mr. As dominant interpersonal concerns cen-

30

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

tered around rejection, shame, and aloneness. He was


able to think about himself and others in complex
ways and to show genuine care and concern toward
other people, although these strengths were often not
manifest because of his interpersonal avoidance. He
had low self-esteem, although he had some intellectual awareness that his feelings toward himself were
unrealistically negative. He often voiced identity concerns, wondering what he was going to do with his life
and where he would fit in and feeling adrift without
either meaningful work or love relationships that
were sustaining. (This is, of course, a very skeletal description of functional domains in Mr. A; for a more
thorough description, and an empirical description
using the Shedler-Westen Assessment Procedure
Q-Sort, see Westen 1998.)

Conclusions
These observations are highly schematic versions of
what an FFM or functional-domains (or psychodynamic) account might offer in describing this case.
Nevertheless, they provide some sense of how one
might conceptualize a case from two very different
theoretical perspectivesnotably a case on which
Axis II would be silent because the patients pathology
is not severe enough for an Axis II diagnosis. Theory,
research, and this brief case example all suggest that
including a broader range of personality pathology
should be one of the primary goals guiding the revision of Axis II in DSM-V.

REFERENCES
Abend S, Porder MS, Willick MS: Borderline Patients: Psychoanalytic Perspectives. Madison, CT, International
Universities Press, 1983
Adler G, Buie D: Aloneness and borderline psychopathology: the possible relevance of child development issues.
Int J Psychoanal 60:8396, 1979
Allport G: Personality: A Psychological Interpretation. New
York, Henry Holt, 1937
Allport G, Odbert H: Trait-names: a psycho-lexical study, in
Psychological Monographs, Vol 47. Princeton, NJ, Psychological Review Co, 1936
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition. Washington,
DC, American Psychiatric Association, 1980
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition. Washington,
DC, American Psychiatric Association, 1994

American Psychiatric Association: Diagnostic and Statistical


Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association,
2000
Aron L: A Meeting of Minds: Mutuality in Psychoanalysis,
Vol 4. New York, The Analytic Press, 1996
Axelrod S, Widiger T, Trull T, et al: Relations of five-factor
model antagonism facets with personality disorder
symptomatology. J Pers Assess 69:297313, 1997
Baker L, Silk KR, Westen D, et al: Malevolence, splitting, and
parental ratings by borderlines. J Nerv Ment Dis 180:258
264, 1992
Balint M: The Basic Fault: Therapeutic Aspects of Regression. Evanston, IL, Northwestern University Press, 1969
Bandura A: Social Foundations of Thought and Action. Englewood Cliffs, NJ, Prentice-Hall, 1986
Bandura A: Social cognitive theory of personality, in Handbook of Personality: Theory and Research, 2nd Edition.
Edited by Pervin L, John O. New York, Guilford, 1999,
pp 154196
Beck A: Cognitive aspects of personality disorders and their
relation to syndromal disorders: a psychoevolutionary
approach, in Personality and Psychopathology. Edited
by Cloninger CR. Washington, DC, American Psychiatric Association, 1999, pp 411429
Beck A, Butler A, Brown G, et al: Dysfunctional beliefs discriminate personality disorders. Behav Res Ther 39:1213
1225, 2001
Beck A, Freeman A, Davis DD: Cognitive Therapy of Personality Disorders, 2nd Edition. New York, Guilford, 2003
Bellak L, Chassan JB, Gediman HK, et al: Ego function assessment of analytic psychotherapy combined with
drug therapy. J Nerv Ment Dis 157:465469, 1973
Benjamin L[S]: Interpersonal Diagnosis and Treatment of
Personality Disorders. New York, Guilford, 1993
Benjamin LS: Interpersonal Diagnosis and Treatment of Personality Disorders, 2nd Edition. New York, Guilford,
1996a
Benjamin LS: An interpersonal theory of personality disorders, in Major Theories of Personality Disorder. Edited
by Clarkin JF, Lenzenweger MF. New York, Guilford,
1996b, pp 141220
Blanck G, Blanck R: Ego Psychology: Theory and Practice.
New York, Columbia University Press, 1974
Bleuler E: Dementia Praecox or the Group of Schizophrenias
(1911). New York, International Universities Press, 1950
Brenner C: The Mind in Conflict. New York, International
Universities Press, 1982
Cadoret RJ, Yates WR, Troughton E, et al: Geneticenvironmental interaction in the genesis of aggressivity and
conduct disorders. Arch Gen Psychiatry 52:916924,
1995
Cantor N, Harlow RE: Personality, strategic behavior, and
daily life problem solving. Current Directions in Psychological Science 3:169172, 1994
Cantor N, Kihlstrom JF: Personality and Social Intelligence.
Englewood Cliffs, NJ, Prentice-Hall, 1987

Theories of Personality and Personality Disorders

Caspi A, McClay J, Moffitt T, et al: Role of genotype in the cycle of violence in maltreated children. Science 297:851
853, 2002
Cloninger CR: The genetics and psychobiology of the sevenfactor model of personality, in Biology of Personality
Disorders. Edited by Silk KR (Review of Psychiatry Series, Vol 17; Oldham JM, Riba MB, series eds). Washington, DC, American Psychiatric Press, 1998, pp 6392
Cloninger RC, Svrakic D: Differentiating normal and deviant
personality by the seven-factor personality model, in Differentiating Normal and Abnormal Personality. Edited
by Strack S, Lorr M. New York, Springer, 1994, pp 4064
Cloninger RC, Svrakic D, Przybeck T: A psychobiological
model of temperament and character. Arch Gen Psychiatry 50:975990, 1993
Coker L, Samuel D, Widiger T: Maladaptive personality functioning within the Big Five and the five-factor model.
J Personal Disord 16:385401, 2002
Costa P, McCrae R: Longitudinal stability of adult personality, in Handbook of Personality Psychology. Edited by
Hogan R, Johnson J. San Diego, CA, Academic Press,
1997, pp 269290
Dahl A: The personality disorders: a critical review of family, twin, and adoption studies. J Personal Disord 7
(suppl):8699, 1993
Damon W, Hart D: Self-Understanding in Childhood and
Adolescence. New York, Cambridge University Press,
1988
Davis R: Millon: Essentials of his science, theory, classification, assessment, and therapy. J Pers Assess 72:330352,
1999
Depue R, Collins P: Neurobiology of the structure of personality: dopamine, facilitation of incentive motivation,
and extraversion. Behav Brain Sci 22:491569, 1999
Depue R, Lenzenweger M: A neurobehavioral dimensional
model, in Handbook of Personality Disorders: Theory,
Research, and Treatment. Edited by Livesley J. New
York, Guilford, 2001, pp 136176
Espelage DL, Mazzeo SE, Sherman R, et al: MCMI-II profiles
of women with eating disorders: a cluster analytic investigation. J Personal Disord 16:453463, 2002
Eysenck H: The Biological Basis of Personality. Springfield,
IL, Charles C Thomas, 1967
Eysenck H: A Model for Personality. New York, SpringerVerlag, 1981
Eysenck HJ: Biological dimensions of personality, in Handbook of Personality: Theory and Research. Edited by
Pervin LA. New York, Guilford, 1990, pp 244276
Fairbairn WR: Psychoanalytic Studies of the Personality.
London, England, Tavistock, 1952
Fonagy P, Target M: Attachment and reflective function:
their role in self-organization. Dev Psychopathol 9:679
700, 1997
Fonagy P, Steele H, Steele M: Maternal representations of attachment during pregnancy predict the organization of
infantmother attachment at one year of age. Child Dev
62:891905, 1991

31

Fonagy P, Gergely G, Jurist EL, et al: Affect Regulation, Mentalization, and the Development of the Self. New York,
Other Press, 2002
Fonagy P, Target M, Gergely G, et al: The developmental
roots of borderline personality disorder in early attachment relationships: a theory and some evidence. Psychoanalytic Inquiry 23:412459, 2003
Goldberg L: The structure of phenotypic personality traits.
Am Psychol 48:2634, 1993
Goldstein W: DSM-III and the narcissistic personality. Am
J Psychother 39:416, 1985
Greenberg JR, Mitchell S: Object Relations in Psychoanalytic
Theory. Cambridge, MA, Harvard University Press,
1983
Gunderson JG: Borderline Personality Disorder: A Clinical
Guide. Washington, DC, American Psychiatric Publishing, 2001
Harter S: The Construction of the Self: A Developmental Perspective. New York, Guilford, 1999
Herpertz SC, Dietrich TM, Wenning B, et al: Evidence of abnormal amygdala functioning in borderline personality
disorder: a functional MRI study. Biol Psychiatry 50:
292298, 2001
Horowitz M: Introduction to Psychodynamics: A Synthesis.
New York, Basic Books, 1988
Horowitz M: Cognitive Psychodynamics: From Conflict to
Character. New York, Wiley, 1998
Kendler K, Walsh D: Schizotypal personality disorder in parents and the risk for schizophrenia in siblings. Schizophr
Bull 21:4752, 1995
Kernberg O: Borderline Conditions and Pathological Narcissism. Northvale, NJ, Jason Aronson, 1975a
Kernberg O: Transference and countertransference in the
treatment of borderline patients. J Natl Assoc Priv Psychiatr Hosp 7:1424, 1975b
Kernberg O: Severe Personality Disorders. New Haven, CT,
Yale University Press, 1984
Kernberg O: A psychoanalytic theory of personality disorders, in Major Theories of Personality Disorder. Edited
by Clarkin J, Lenzenweger M. New York, Guilford,
1996, pp 106140
Kernberg O: Pathological narcissism and narcissistic personality disorder: theoretical background and diagnostic
classification, in Disorders of Narcissism: Diagnostic,
Clinical, and Empirical Implications. Edited by Ronningstam E. Washington, DC, American Psychiatric Association, 1998, pp 2951
Kiesler D: The 1982 interpersonal circle: a taxonomy for
complementarity in human transactions. Psychol Rev
90:185214, 1983
Kohut H: The Analysis of the Self: A Systematic Approach to
the Treatment of Narcissistic Personality Disorders.
New York, International Universities Press, 1971
Kohut H: The Restoration of the Self. Madison, WI, International Universities Press, 1977
Kraepelin E: Dementia Praecox and Paraphrenia (1896). Chicago, IL, Chicago Medical Book Co, 1919

32

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Kristensen H, Torgersen S: MCMI-II personality traits and


symptom traits in parents of children with selective mutism: a case-control study. J Abnorm Psychol 110:648652,
2001
Krueger RF: The structure of common mental disorders.
Arch Gen Psychiatry 56:921926, 1999
Leary T: Interpersonal Diagnosis of Personality: A Functional Theory and Methodology for Personality Evaluation. Oxford, England, Ronald Press, 1957
Lenzenweger M: Schizotypy and schizotypic psychopathology, in Origins and Development of Schizophrenia: Advances in Experimental Psychopathology. Edited by
Lenzenweger M, Dworkin R. Washington, DC, American Psychological Association, 1998, pp 93122
Linehan M: Cognitive-Behavioral Treatment of Borderline
Personality Disorder. New York, Guilford, 1993a
Linehan M: Skills-Training Manual for Treatment of Borderline Personality Disorder. New York, Guilford, 1993b
Livesley WJ, Bromley DB: Person Perception in Childhood
and Adolescence. London, England, Wiley, 1973
Livesley WJ, Jang K, Jackson D, et al: Genetic and environmental contributions to dimensions of personality disorder. Am J Psychiatry 150:18261831, 1993
Livesley WJ, Jang KL, Vernon PA: Phenotypic and genetic
structure of traits delineating personality disorder. Arch
Gen Psychiatry 55:941948, 1998
Livesley WJ, Jang K, Vernon P: Genetic basis of personality
structure, in Handbook of Psychology: Personality and
Social Psychology, Vol 5. Edited by Millon T, Lerner M.
New York, Wiley, 2003, pp 5983
Lynam DR, Widiger TA: Using the five-factor model to represent the DSM-IV personality disorders: an expert consensus approach. J Abnorm Psychol 110:401412, 2001
Lyons M, Toomey R, Faraone S, et al: Correlates of psychosis
proneness in relatives of schizophrenic patients. Journal
of Abnormal Psychology and Psychotherapy: Theory,
Research and Practice 104:390394, 1995
Main M: Recent studies in attachment: overview, with selected implications for clinical work, in Attachment
Theory: Social, Developmental, and Clinical Perspectives. Edited by Goldberg S, Muir R, Kerr J. Hillsdale,
NJ, Analytic Press, 1995, pp 407474
Masterson J: Psychotherapy of the Borderline Adult: A Developmental Approach. New York, Brunner/Mazel,
1976
McCrae R, Costa P: Personality in Adulthood. New York,
Guilford, 1990
McCrae R, Costa PL: Personality trait structure as a human
universal. Am Psychol 52:509516, 1997
McWilliams N: Relationship, subjectivity, and inference in
diagnosis, in Making Diagnosis Meaningful: Enhancing
Evaluation and Treatment of Psychological Disorders.
Edited by Barron JW. Washington, DC, American Psychological Association, 1998, pp 197226
Millon T: Modern Psychopathology: A Biosocial Approach
to Maladaptive Learning and Functioning. Philadelphia, PA, WB Saunders, 1969

Millon T: Toward a New Psychology. New York, Wiley, 1990


Millon T, Davis R: Conceptions of personality disorders: historical perspectives, the DSMs, and future directions, in
The DSM-IV Personality Disorders: Diagnosis and
Treatment of Mental Disorders. Edited by Livesley WJ.
New York, Guilford, 1995, pp 328
Millon T, Davis R: An evolutionary theory of personality disorders, in Major Theories of Personality Disorder. Edited by Clarkin J, Lenzenweger M. New York, Guilford,
1996, pp 221346
Millon T, Davis R: The MCMI-III: present and future directions. J Pers Assess 68:6985, 1997
Mischel W: Toward a cognitive social learning reconceptualization of personality. Psychol Rev 39:351364, 1973
Mischel W: On the interface of cognition and personality: beyond the person-situation debate. Am Psychol 34:740
754, 1979
Mischel W: Personality dispositions revisited and revised: a
view after three decades, in Handbook of Personality:
Theory and Research. Edited by Pervin L. New York,
Guilford, 1990, pp 111134
Mischel W, Shoda Y: A cognitive-affective system theory of
personality: reconceptualizing situations, dispositions,
dynamics, and invariance in personality structure. Psychol Rev 102:246268, 1995
Mitchell SA: Relational Concepts in Psychoanalysis: An Integration. Cambridge, MA, Harvard University Press, 1988
Morey L, Gunderson J, Quigley B, et al: The representation
of borderline, avoidant, obsessive-compulsive, and
schizotypal personality disorders by the five-factor
model. J Personal Disord 16:215234, 2002
New A, Siever L: Neurobiology and genetics of borderline
personality disorder. Psychiatric Annals of Clinical Psychiatry 32:329336, 2002
Nigg JT, Goldsmith H: Genetics of personality disorders:
perspectives from personality and psychopathology research. Psychol Bull 115:346380, 1994
Plomin R, Caspi A: Behavioral Genetics and Personality.
New York, Guilford, 1999
Pretzer JL, Beck AT: A cognitive theory of personality disorders, in Major Theories of Personality Disorder. Edited
by Clarkin J, Lenzenweger M. New York, Guilford,
1996, pp 36105
Redl F, Wineman D: Children Who Hate: The Disorganization and Breakdown of Behavior Controls. Glencoe, IL,
Free Press, 1951
Reich J: Familiality of DSM-III dramatic and anxious personality clusters. J Nerv Ment Dis 177:96100, 1989
Reich W: Character Analysis, 3rd Edition (1933). New York,
Simon and Schuster, 1978
Ross S, Lutz C, Bailley S: Positive and negative symptoms of
schizotypy and the five-factor model: a domain and
facet level analysis. J Pers Assess 79:5372, 2002
Schaefer E: Configurational analysis of childrens reports of
parent behavior. J Consult Psychol 29:552557, 1965
Siever L, Davis K: A psychobiological perspective on the personality disorders. Am J Psychiatry 148:16471658, 1991

Theories of Personality and Personality Disorders

Siever L, Torgersen S, Gunderson J, et al: The borderline diagnosis III: identifying endophenotypes for genetic
studies. Biol Psychiatry 51:964968, 2003
Skodol A, Siever L, Livesley W, et al: The borderline diagnosis II: biology, genetics, and clinical course. Biol Psychiatry 51:951963, 2002
Spitzer RL, Endicott J, Gibbon M: Crossing the border into
borderline personality and borderline schizophrenia.
Arch Gen Psychiatry 36:1724, 1979
Sullivan HS: The Interpersonal Theory of Psychiatry. New
York, Norton, 1953
Torgersen S: Genetic and nosological aspects of schizotypal
and borderline personality disorders. Arch Gen Psychiatry 41:546554, 1984
Torgersen S, Lygren S, Oien PA, et al: A twin study of personality disorders. Compr Psychiatry 41:416425, 2000
Trull T, Widiger T, Burr R: A structured interview for the assessment of the Five-Factor Model of personality: facetlevel relations to the axis II personality disorders. J Pers
69:175198, 2001
Turkheimer E, Waldron M: Nonshared environment: a theoretical, methodological, and quantitative review. Psychol Bull 126:78108, 2000
Turkheimer E, Haley A, Waldron M, et al: Socioeconomic
status modifies heritability of IQ in young children.
Psychol Sci 14:623628, 2003
Tyrer P, Johnson T: Establishing the severity of personality
disorder. Am J Psychiatry 153:15931597, 1996
Wachtel P: Psychoanalysis and Behavior Therapy. New
York, Basic Books, 1977
Wachtel P: Psychoanalysis, Behavior Therapy, and the Relational World. Washington, DC, American Psychological
Association, 1997
Westen D: The relations among narcissism, egocentrism,
self-concept, and self-esteem: experimental, clinical and
theoretical considerations. Psychoanalysis and Contemporary Thought 13:183239, 1990a
Westen D: Towards a revised theory of borderline object relations: contributions of empirical research. Int J Psychoanal 71:661693, 1990b
Westen D: Cognitive-behavioral interventions in the psychoanalytic psychotherapy of borderline personality disorders. Clin Psychol Rev 11:211230, 1991a
Westen D: Social cognition and object relations. Psychol Bull
109:429455, 1991b
Westen D: Toward an integrative model of affect regulation:
applications to social-psychological research. J Pers
62:641667, 1994
Westen D: A clinical-empirical model of personality: life after the Mischelian ice age and the NEO-lithic era. J Pers
63:495524, 1995

33

Westen D: A model and a method for uncovering the nomothetic from the idiographic: an alternative to the fivefactor model? J Res Pers 30:400413, 1996
Westen D: Case formulation and personality diagnosis: two
processes or one? In Making Diagnosis Meaningful: Enhancing Evaluation and Treatment of Psychological
Disorders. Edited by Barron JW. Washington, DC,
American Psychological Association, 1998, pp 111138
Westen D, Muderrisoglu S: Assessing personality disorders
using a systematic clinical interview: evaluation of an
alternative to structured interviews. J Personal Disord
17:351369, 2003
Westen D, Shedler J: Revising and assessing Axis II, part 1:
developing a clinically and empirically valid assessment method. Am J Psychiatry 156:258272, 1999a
Westen D, Shedler J: Revising and assessing Axis II, part 2:
toward an empirically based and clinically useful classification of personality disorders. Am J Psychiatry
156:273285, 1999b
Widiger T: Personality disorders in the 21st century. J Personal Disord 14:316, 2000
Widiger T, Costa P: Personality and personality disorders.
Journal of Abnormal Psychology and Psychotherapy:
Theory, Research, and Practice 103:7891, 1994
Widiger T, Frances A: Towards a dimensional model for the
personality disorders, in Personality Disorders and the
Five-Factor Model of Personality. Edited by Costa P, Widiger T. Washington, DC, American Psychological Association, 1994, pp 1939
Widiger T, Trull TJ, Clarkin JF, et al: A description of the
DSM-IV personality disorders with the five-factor
model of personality, in Personality Disorders and the
Five-Factor Model of Personality, 2nd Edition. Washington, DC, American Psychological Association, 2002,
pp 8999
Wolf E: Treating the Self: Elements of Clinical Self-Psychology. New York, Guilford, 1988
Young J: Cognitive Therapy for Personality Disorders: A
Schema-Focused Approach. Sarasota, FL, Professional
Resource Exchange, 1990
Young J, Gluhoski V: Schema-focused diagnosis for personality disorders, in Handbook of Relational Diagnosis and
Dysfunctional Family Patterns. Edited by Kaslow F. Oxford, England, Wiley, 1996, pp 300321
Young J, Lindemann M: An integrative schema-focused
model for personality disorders, in Clinical Advances in
Cognitive Psychotherapy: Theory and Application. Edited by Leahy R, Dowd T. New York, Springer, 2002, pp
93109
Young J, Klosko J, Weishaar M: Schema therapy for borderline personality disorder, in Schema Therapy: A Practitioners Guide. New York, Guilford, 2003

This page intentionally left blank

3
Categorical and Dimensional
Models of Personality
Disorders
Thomas A. Widiger, Ph.D.
Stephanie N. Mullins-Sweatt, M.A.

personality disorder. The purpose of this chapter is to


provide the rationale and empirical support for this
perspective and to indicate how personality disorders
could be conceptualized as maladaptive variants of
continuously distributed personality traits.

CATEGORICAL AND DIMENSIONAL


MODELS OF PERSONALITY DISORDERS
The conceptualization of personality disorders in the
American Psychiatric Associations Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR; American Psychiatric Association 2000) represents the categorical perspective that [p]ersonality [d]isorders are
qualitatively distinct clinical syndromes (p. 689).
Nevertheless, it is also acknowledged that an alternative to the categorical approach is the dimensional
perspective that [p]ersonality [d]isorders represent
maladaptive variants of personality traits that merge
imperceptibly into normality and into one another
(p. 689). As concluded by a joint committee of the
American Psychiatric Association and the National Institute of Mental Health addressing issues and proposals for DSM-V, there is a clear need for dimensional models to be developed and for their utility to
be compared with that of existing typologies (Rounsaville et al. 2002, p. 12). The committee emphasized in
particular the development of a dimensional model of

LIMITATIONS OF THE
CATEGORICAL MODEL
Four concerns commonly cited with respect to the categorical model of personality disorder diagnosis are excessive diagnostic co-occurrence, heterogeneity among
persons with the same diagnosis, absence of a nonarbitrary boundary with normal functioning, and inadequate coverage of maladaptive personality functioning.
Each of these concerns is discussed briefly in turn.

Excessive Diagnostic Co-Occurrence


DSM-IV-TR provides diagnostic criteria sets to help
guide the clinician toward the correct diagnosis and a
section devoted to differential diagnosis that indicates
35

36

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

how to differentiate [the] disorder from other disorders


that have similar presenting characteristics (American
Psychiatric Association 2000, p. 10). The intention of this
information is to help the clinician determine which particular disorder is present, the selection of which would
ideally indicate the presence of a specific pathology that
will explain the occurrence of the symptoms and suggest a specific treatment to ameliorate the patients suffering (Frances et al. 1995).
It is evident, however, that DSM-IV-TR routinely
fails in the goal of guiding the clinician to the presence
of one specific personality disorder. A number of reviews have indicated that many patients meet diagnostic criteria for an excessive number of personality
disorder diagnoses (Bornstein 1998; Lilienfeld et al.
1994; Livesley 2003; Oldham et al. 1992; Widiger and
Trull 1998). Thus, the maladaptive personality functioning of patients does not appear to be adequately
described by a single diagnostic category. No person is
generally well described by just one word. Each person is more accurately described by a constellation of
personality traits (John and Srivastava 1999).
One approach to diagnostic co-occurrence is to implement hierarchical decision rules. Hierarchical decision rules would eliminate the occurrence of multiple
diagnoses, and they may be consistent with how personality disorders are diagnosed in clinical practice
(Gunderson 1992). Clinicians generally provide only
one personality disorder diagnosis per patient, possibly using their own decision rules for which diagnosis
takes precedence (Herkov and Blashfield 1995; Zimmerman and Mattia 1999). However, one limitation of
a hierarchical decision rule is the difficulty of establishing a compelling rationale for which diagnosis
should take precedence (Gunderson 1992). In addition, any such rule would not actually make the comorbidity go away. For example, borderline patients
with obsessive-compulsive personality traits will still
have obsessive-compulsive personality traits even if
those traits are not included in the diagnosis (Zimmerman and Mattia 1999).

Heterogeneity Among Persons With


the Same Diagnosis
There are also important differences among the persons who share the same personality disorder diagnosis. For example, patients with the same diagnosis will
vary substantially with respect to which diagnostic
criteria were used to make the diagnosis (Clark 1992;
Shea 1992), and the differences are not trivial (Millon
et al. 1996). For example, only a subset of persons who

meet the DSM-IV-TR criteria for antisocial personality


disorder will have the prototypic features of the callous, ruthless, arrogant, charming, and scheming psychopath (Hare et al. 1991), and there are even important differences among the persons who would be
diagnosed as psychopathic (Brinkley et al. 2004). One
common distinction is between the successful and unsuccessful psychopath, with the former having high
levels of diligence, competence, and achievementstriving, whereas the latter is characterized by a laxness, irresponsibility, and negligence (Lynam 2002).
Similar distinctions are made for other personality
disorders (Millon et al. 1996), such as the differentiation of borderline psychopathology with respect to the
dimensions of affective dysregulation, impulsivity,
and behavioral disturbance (Sanislow et al. 2002), and
the differentiation of dependent personality disorder
into submissive, exploitable, and affectionate variants
(Pincus and Wilson 2001).

Inconsistent, Unstable, and Arbitrary


Diagnostic Boundaries
An additional limitation of the categorical model is the
difficulty of establishing a nonarbitrary boundary between disordered and normal personality functioning.
One of the innovations of DSM-III (American Psychiatric Association 1980) was the provision of explicit diagnostic criteria, including a specified threshold for a disorders diagnosis. However, the diagnostic thresholds
lack a compelling rationale (Tyrer and Johnson 1996). In
fact, no explanation or justification has ever been provided for most of them (Widiger and Corbitt 1994).
The thresholds for the DSM-III schizotypal and
borderline diagnoses are the only two for which a rationale has been provided. The DSM-III requirements
that the patient have four of eight features for the
schizotypal and five of eight for the borderline diagnosis were determined on the basis of maximizing
agreement with similar diagnoses provided by clinicians (Spitzer et al. 1979). However, the current diagnostic thresholds for these personality disorders bear
little resemblance to the original thresholds established for DSM-III. Blashfield et al. (1992) reported a
kappa of only -0.025 for the DSM-III and DSM-III-R
(American Psychiatric Association 1987) schizotypal
personality disorders, with a reduction in prevalence
from 11% to 1%. Seemingly minor changes to diagnostic criteria sets have resulted in unexpected and substantial shifts in prevalence rates that profoundly complicate scientific theory and public health decisions
(Blashfield et al. 1992; Narrow et al. 2002).

Categorical and Dimensional Models of Personality Disorders

Inadequate Coverage
In addition to the problem of excessive diagnostic cooccurrence, there is the opposite problem of inadequate coverage. Clinicians provide a diagnosis of personality disorder not otherwise specified (NOS) when
they determine that a person has a personality disorder that is not adequately represented by any one of
the 10 officially recognized diagnoses (American Psychiatric Association 2000). Personality disorder NOS
is often the single most frequently used diagnosis in
clinical practice; one explanation for this is that the existing categories are not providing adequate coverage
(Verheul and Widiger 2004). Westen and ArkowitzWesten (1998) surveyed 238 psychiatrists and psychologists with respect to their clinical practice and reported that the majority of patients with personality
pathology significant enough to warrant clinical psychotherapeutic attention (60.6%) are currently undiagnosable on Axis II (p. 1769). The clinicians reported
personality traits concerning commitment, intimacy,
shyness, work inhibition, perfectionism, and devaluation of others that were not well described by any of
the existing diagnostic categories.
One approach to this problem is to add more diagnostic categories, but there is considerable reluctance
to do so, in part because adding categories would increase further the difficulties with excessive diagnostic co-occurrence and differential diagnosis (Pincus et
al. 2003). A dimensional model that is reasonably comprehensive would be able to cover a greater range of
maladaptive personality functioning without requiring additional diagnostic categories: by avoiding the
inclusion of redundant, overlapping diagnoses; by organizing the traits within a hierarchical structure; by
representing a broader range of maladaptive personality functioning along a single dimension; and by allowing for the representation of relatively unique or
atypical personality profiles (Samuel and Widiger
2004).

VALIDITY OF DIMENSIONAL AND


CATEGORICAL MODELS
A variety of statistical and methodological approaches
for addressing the validity of categorical and dimensional models of classification have been used, including (but not limited to) the search for evidence of incremental validity, bimodality, discrete breaks within
distributions, and reproducibility of factor analytic solutions across groups; as well as latent class, item re-

37

sponse, taxometric, and admixture analyses (Haslam


2003; Klein and Riso 1993).
Support for a dimensional model is provided in
part by the finding that the maladaptive personality
traits included within the diagnostic criteria for the
DSM-IV-TR personality disorders are present within
members of the general population who would not be
diagnosed with a DSM-IV-TR personality disorder.
For example, much (if not all) of the fundamental
symptomatology of the DSM-IV-TR personality disorders can be understood as maladaptive variants of
personality traits included within general models of
personality functioning (Saulsman and Page 2004; Widiger and Costa 2002). The symptoms of borderline
personality disorder (BPD) can be understood as extreme variants of the angry hostility, vulnerability,
anxiousness, depressiveness, and impulsivity included within the broad domain of neuroticism (identified by others as negative affectivity or emotional instability) that is evident within the general population
(Clarkin et al. 1993; Morey and Zanarini 2000; Trull et
al. 2003). Similarly, much of the symptomatology of
antisocial personality disorder appears to be an extreme variant of low conscientiousness (rashness, negligence, hedonism, immorality, undependability, and
irresponsibility) and high antagonism (manipulativeness, deceptiveness, exploitativeness, aggressiveness,
callousness, and ruthlessness) that have long been
evident within the general population (Miller and
Lynam 2003; Miller et al. 2001).
Over 50 published studies have suggested that the
personality disorders included within DSM-IV appear
to be maladaptive variants of common personality
traits identified within the general population (Widiger and Costa 2002). Trull et al. (2003) demonstrated
that the extent to which a persons personality trait
profile matched the profile of a prototypic case of BPD
correlated as highly with measures of BPD as measures of BPD correlated with one another, and that this
general personality trait index of BPD replicated the
relationship of the clinical measures with external validators. Miller and Lynam (2003) demonstrated similarly that a general personality measure of psychopathy predicted drug usage, delinquency, risky sexual
behavior, and aggression; as well as several laboratory
assessments of pathologies hypothesized to underlie
the personality disorder of psychopathy, including
willingness to delay gratification in a time-discounting task and a preference for aggressive responses in a
social-information processing paradigm.
The structure and heritability of personality disorder
symptomatology within general community samples of

38

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

persons without DSM-IV-TR personality disorders is


convergent with the structure and heritability observed
among persons who have been diagnosed with these
disorders (Tyrer and Alexander 1979). Livesley et al.
(1998) compared the phenotypic and genetic structure of
a comprehensive set of personality disorder symptoms
in samples of 656 patients with personality disorder, 939
general community participants, and 686 twin pairs.
Principal components analysis yielded four broad dimensions (emotional dysregulation, dissocial behavior,
inhibitedness, and compulsivity) that were replicated
across all three samples. Multivariate genetic analyses
also yielded the same four factors. The stable structure
of traits across clinical and nonclinical samples is consistent with dimensional representations of personality disorders (Livesley et al. 1998, p. 941). Livesley and colleagues also noted the remarkable consistency of the four
broad domains of personality disorder with four of the
five broad domains consistently identified in studies of
general personality functioning. They concluded that
the higher-order traits of personality disorder strongly
resemble dimensions of normal personality (p. 941).
Joint factor analyses of measures of the general personality functioning and comprehensive representations of personality disorder symptomatology have
consistently confirmed a common underlying structure
(Cannon et al. 2003; Clark and Livesley 2002). In sum, it
is striking that an extensive history of research to develop a dimensional model of normal personality functioning that has been confined to community populations is so closely congruent with a model that was
derived from an analysis confined to personality disorder symptoms (Widiger 1998, p. 865). OConnor (2002)
submitted the correlation and factor loading matrices
for 37 psychopathology and personality inventories obtained from multiple data sets to round-robin confirmatory factor analyses to determine whether there are differences in the dimensional structure between clinical
and nonclinical respondents. He reported quite consistent evidence for high levels of similarity between the
normal and abnormal populations with respect both to
the number of factors and the factor patterns. OConnor
concluded that the dimensional universes of normality and abnormality are apparently the same, at least according to data derived from contemporary assessment
instruments (p. 962).

ALTERNATIVE MODELS
The limitations of the categorical model are becoming
increasingly recognized by theorists, researchers, and

clinicians (Oldham and Skodol 2000; Rounsaville et al.


2002). An expected response to this recognition is the
development of proposals for alternative dimensional
models. Quite a few dimensional models of personality disorder have been developed; however, space limitations prohibit a comprehensive summary of all of
them. We describe in this section several alternative
strategies for developing a dimensional model of personality disorder. Additional models beyond those described herein are 1) Eysencks (1987) three dimensions
of neuroticism, extraversion, and psychoticism; 2) the
Personality PsychopathologyFive (PSY-5), consisting
of positive emotionality/extraversion, aggressiveness,
constraint, negative emotionality/neuroticism, and
psychoticism (Harkness et al. 1995); 3) the three selfother, pleasure-pain, and active-passive polarities hypothesized by Millon et al. (1996) and assessed by the
Millon Index of Personality Styles (MIPS, Millon 1994);
4) Tyrers (1988) four antisocial, dependent, inhibited,
and withdrawn dimensions of personality disorder;
and 5) Zuckermans (2002) five dimensions of sociability, activity, aggression-hostility, impulsive sensationseeking, and neuroticism-anxiety.

Dimensional Profile of Personality Disorder


Diagnostic Categories
A straightforward approach that would involve the
least amount of disruption to the existing nomenclature
is to provide a dimensional profile of maladaptive personality functioning in terms of the existing (or somewhat revised) diagnostic categories (Oldham and
Skodol 2000; Tyrer and Johnson 1996; Widiger and
Sanderson 1995). A personality disorder could be characterized as prototypic if all of the diagnostic criteria
are met, moderately present if one or two criteria beyond the threshold for a categorical diagnosis are
present, threshold if the patient just barely meets diagnostic threshold, subthreshold if symptoms are present
but are just below diagnostic threshold, traits if one to
three symptoms are present, and absent if no diagnostic
criteria are present (Oldham and Skodol 2000). Oldham
and Skodol (2000) proposed further that if a patient
meets diagnostic criteria for three or more personality
disorders, then a diagnosis of extensive personality
disorder could be provided, along with an indication
of the extent to which each personality disorder is
present.
Westen and Shedlers (2000) prototypal matching
proposal is similar to the proposal of Oldham and
Skodol (2000) in that it retains the existing or at least
somewhat revised diagnostic categories, each of which

Categorical and Dimensional Models of Personality Disorders

would be rated on a five-point scale. However, an important difference is that this five-point rating would
not be based on the number of diagnostic criteria.
Shedler and Westen (2004) suggested that specific and
explicit diagnostic criteria sets are impractical and unnecessary in clinical practice. They proposed instead
that the diagnostic manual provide a narrative description of a prototypic case of each personality disorder,
with the clinician indicating on a five-point scale the extent to which the actual case matches this description
(i.e., 1=description does not apply; 2=only minor features; 3= significant features; 4=strong match, patient
has the disorder; and 5=exemplifies the disorder, prototypic case). An additional distinction is that the narrative descriptions would not be confined to the eight or
nine diagnostic criteria currently provided but could instead be expanded to provide more extensive descriptions of prototypic cases. Shedler and Westen (2004)
provide descriptions of each personality disorder using
the Shedler-Westen Assessment Procedure-200 (SWAP200). The SWAP-200 includes 200 diagnostic criteria
(approximately half of which are taken from DSM-IVTR), drawn from the psychoanalytic and wider personality disorder literature (Shedler 2002).

Dimensional Reorganization of
Personality Disorder Symptoms
The proposals of Oldham and Skodol (2000), Tyrer
and Johnson (1996), and Westen and Shedler (2000)
would largely retain the existing personality disorder
categories but provide a means for how each could be
described in a more quantitative manner. A potential
limitation of these proposals is that there might be underlying dimensions of maladaptive personality functioning that cut across the existing diagnostic constructs, contributing to their diagnostic co-occurrence.
The proposals of Livesley (2003) and Clark (1993) are
efforts to identify these underlying dimensions of
maladaptive personality functioning.
Livesley (2003) approached the development of a
dimensional model of personality disorders empirically. He obtained personality disorder symptoms and
features from a thorough content analysis of the personality disorder literature. An initial list of criteria
was then coded by clinicians with respect to their prototypicality for respective personality disorders. One
hundred scales (each with 16 items) were submitted to
a series of factor analyses to derive a set of 18 fundamental dimensions of personality disorder that cut
across the existing diagnostic categories (e.g., anxiousness, self-harm, intimacy problems, social avoidance,

39

passive opposition, and interpersonal disesteem). Additional analyses indicate that these 18 dimensions can
be subsumed within four higher-order dimensions:
emotional dysregulation, dissocial behavior, inhibitedness, and compulsivity. Assessment of the 18-factor
model has been provided by the self-report Dimensional Assessment of Personality PathologyBasic
Questionnaire (DAPP-BQ; Livesley 2003).
Clarks (1993) approach was quite similar to that of
Livesley (2003). The DSM-III-R personality disorder
criteria, along with items obtained from the broader
personality disorder literature and selected Axis I disorders (i.e., traitlike manifestations of anxiety and
mood disorders), were sorted by clinicians into 22
conceptually similar symptom clusters. Factor analyses of these 22 symptom clusters yielded 12 dimensions of maladaptive personality functioning (e.g.,
self-harm, entitlement, eccentric perceptions, workaholism, detachment, and manipulation). These 12 dimensions of abnormal personality functioning are related conceptually to three higher-order factors of
general personality hypothesized by Watson and colleagues (1999): negative affectivity, positive affectivity, and constraint. Assessment of the 12-factor model
has been provided by the self-report Schedule for
Nonadaptive and Adaptive Personality (SNAP; Clark
1993). However, support for the three-factor structure
of the SNAP has been provided by other instruments,
particularly the Multidimensional Personality Questionnaire (MPQ; Tellegen in press). The MPQ has alternative subscales for the three broad domains, including stress reaction, alienation, and aggression
within negative emotionality; control, traditionalism,
and harm avoidance within constraint; and achievement, social closeness, social potency, and well-being
within positive emotionality, and it also has an additional scale of absorption.

Clinical Spectra Models


Clark (1993) included within her factor analyses of personality disorder symptoms traitlike manifestations of
anxiety and mood disorders because the diagnostic cooccurrence of personality and Axis I disorders could be
due to the presence of common underlying dimensions of maladaptive personality functioning (i.e., temperaments of negative affectivity, positive affectivity,
and constraint; Clark and Watson 1999). A proposal by
Siever and Davis (1991) was concerned specifically
with the diagnostic co-occurrence of the personality
and Axis I disorders. The authors suggested that there
is no meaningful boundary between the personality

40

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

and Axis I disorders and proposed that personality


and other mental disorders be collapsed into four
broad clinical spectra consisting of cognitive/perceptual organization, impulsivity/aggression, affective
instability, and anxiety/inhibition.
A suggestion of the clinical spectra model is to
reformulate most of the existing personality disorders
as early onset, chronic variants of an existing Axis I
disorder (First et al. 2002; Siever and Davis 1991).
Avoidant personality disorder could be replaced by
generalized social phobia; depressive personality disorder by early onset dysthymia; BPD by an affective
dysregulation disorder; schizotypal and schizoid personality disorders by an early onset and chronic variant of schizophrenic pathology (as schizotypal is already classified in ICD-10; World Health Organization
1992); paranoid personality disorder by an early onset,
chronic, and milder variant of a delusional disorder;
obsessive-compulsive personality disorder by a generalized and chronic variant of obsessive-compulsive
anxiety disorder; and antisocial personality disorder
by an adult variant of conduct (disruptive behavior)
disorder. This reformulation would leave just four
personality disorders unaccounted for (i.e., histrionic,
narcissistic, dependent, and passive-aggressive) that
could then be deleted from the manual as falling outside of the existing clinical spectra.
There is little direct empirical support for the four
clinical spectra proposal of Siever and Davis (1991),
due in part to the absence of an instrument for its assessment. Nevertheless, there is substantial empirical
support for the existence of two fundamental dimensions of internalization and externalization that cut
across the Axes I and II division (Krueger 1999; Krueger and Tackett 2003). The internalization and externalization dimensions identified by Krueger and colleagues do not map perfectly onto the four clinical
spectra of Siever and Davis, but it is apparent that the
spectra of affective instability and anxiety/inhibition
could be folded into the domain of internalization,
and that of impulsivity/aggression into the domain of
externalization.

Dimensional Models of General


Personality Functioning
Personality disorders may not only be on a continuum
with Axis I disorders, they may also be on a continuum
with general personality functioning, contributing to
the absence of a clear boundary between normal and
abnormal personality functioning and to the presence
of a considerable amount of personality disorder symp-

tomatology within the general population (Livesley


2003; Widiger and Sanderson 1995). As indicated earlier, the 12 personality disorder scales of the SNAP are
related conceptually to the three-factor model of general personality functioning proposed by Watson et al.
(1999).

Five-Factor Model
An additional model of general personality functioning is the five-factor model (FFM), derived originally
from factor analytic studies of extensive samples of
trait terms within the English language (John and
Srivastava 1999). In the FFM, the relative importance
of a trait is indicated by the number of terms that have
been developed within a language to describe the various degrees and nuances of that trait, and the structure of the trait is evident by the relationship among
the trait terms (Goldberg 1993). This lexical approach
to personality description has emphasized five broad
domains of personality, presented in their order of
importance as extraversion (or surgency) versus introversion; agreeableness versus antagonism; conscientiousness; emotional instability (or neuroticism); and
unconventionality (or openness). The five broad domains have been replicated in lexical studies of the
trait terms in a wide variety of other languages, including Czech, Dutch, French, German, Hungarian,
Italian, Korean, and Polish, although this research has
also suggested that an additional, smaller factor may
also emergehonesty-humilitythat is currently included largely as a component of agreeableness (Ashton et al. 2004). Each of the five broad domains has
been further differentiated by Costa and McCrae
(1992) into more specific facets. For example, the facets
of agreeableness versus antagonism are trust versus
mistrust, straightforwardness versus deception, altruism versus exploitation, compliance versus opposition, modesty versus arrogance, and tender-mindedness versus tough-mindedness.
The FFM is the predominant model in general personality research, with extensive applications in the
fields of health psychology, aging, and developmental
psychology (McCrae and Costa 1999). Empirical support for the FFM is extensive, including convergent
and discriminant validity at both the domain and facet
levels across self, peer, and spouse ratings; temporal
stability across 710 years; and heritability (McCrae
and Costa 1999; Plomin and Caspi 1999); as well as
links to a wide variety of important life outcomes,
such as mental health (Basic Behavioral Science Task
Force of the National Advisory Mental Health Council
1996), career success (Judge et al. 1999), and mortality

Categorical and Dimensional Models of Personality Disorders

(Friedman et al. 1995). Adaptive and maladaptive


variants of each of the two poles of the 30 facets have
been described (Widiger et al. 2002), and descriptions
by researchers (Lynam and Widiger 2001) and by clinicians (Samuel and Widiger 2004) of each of the DSMIV-TR personality disorders in terms of the FFM have
been provided. A number of alternative measures of
the FFM have been developed. The most commonly
used self-report measure is the NEO Personality
InventoryRevised (NEO-PI-R, Costa and McCrae
1992); a semistructured interview that includes the
maladaptive variants of each pole of each facet was
developed by Trull et al. (1998).

Interpersonal Circumplex
Some theoretical models of personality disorders suggest that they are essentially, if not entirely, disorders
of interpersonal relatedness (Benjamin 1996; Kiesler
1996). All forms of normal and abnormal interpersonal relatedness can be well described as some combination of two fundamental dimensions, identified
by Wiggins (2003) as agency (dominance versus submission) and communion (affiliation, or love versus
hate). Dependent personality disorder, for example,
would represent maladaptively extreme levels of submissiveness and affiliation (Pincus and Wilson 2001).
There are a number of different self-report measures of this interpersonal circumplex (IPC) (Wiggins
2003), with the most popular being perhaps the Interpersonal Adjective ScaleBig Five Version (which includes three additional scales to provide a joint assessment of the FFM and the IPC; Wiggins 2003). The
Wisconsin Personality Disorders Inventory (Klein et
al. 1993) is a self-report inventory for the assessment of
the DSM-IV personality disorders from the perspective of the IPC. Compelling empirical support has
been obtained for an IPC understanding of many of
the personality disorders (Kiesler 1996), particularly
dependent, schizoid, avoidant, histrionic, and passive-aggressive, although this research has also suggested that some aspects of other personality disorders are not well accounted for by the IPC, such as the
affective dysregulation of BPD, the impulsivity of antisocial personality, and the workaholism of obsessivecompulsive disorder (Widiger and Hagemoser 1997).

41

tal dimensions of personality based on a synthesis of


information from family studies, studies of longitudinal development, and psychometric studies of personality structure, as well as neuropharmacologic and
neuroanatomical studies of behavioral conditioning
and learning in man and other animals (p. 574). The
three dimensions were novelty seeking (behavioral activation: exhilaration or excitement in response to
novel stimuli or cues for potential rewards or potential
relief from punishment); harm avoidance (behavioral
inhibition: intense response to signals of aversive
stimuli); and reward dependence (behavioral maintenance: response to signals of reward or to resist extinction of behavior that has been previously reinforced).
Each was hypothesized to be associated with a particular monoamine neuromodulator (i.e., dopamine, serotonin, and norepinephrine, respectively). The theory was revised subsequently to include four rather
than three temperaments (persistence was separated
from reward dependence), along with three additional character dimensions.
The four temperaments reflect innate dispositions
to respond to stimuli in a consistent manner; the character dimensions are considered to be individual differences that develop through a nonlinear interaction
of temperament, family environment, and life experiences (Svrakic et al. 2002). The three character dimensions are self-directedness (responsible, goal-directed
vs. insecure, inept); cooperativeness (helpful, empathic vs. hostile, aggressive); and self-transcendence
(imaginative, unconventional vs. controlling, materialistic). The presence of a personality disorder is indicated by low levels of cooperativeness, self-transcendence, and, most importantly, self-directedness (the
ability to control, regulate, and adapt behavior); and
the specific variants of personality disorder are governed by the four temperaments (Cloninger 2000). The
seven factors (four temperament and three character
dimensions) are assessed by the self-report Temperament and Character Inventory (TCI; Cloninger 2000).
Extensive research concerning Cloninger s sevenfactor model is detailed within his chapter in this text
(Chapter 9, Genetics) and elsewhere (Cloninger
1998; Cloninger and Svrakic 1999).

Seven-Factor Model of Cloninger

INTEGRATION OF ALTERNATIVE MODELS

Cloninger (2000) also developed a dimensional model


of general personality functioning that would include
both normal and abnormal personality traits. He originally hypothesized the existence of three fundamen-

There are notable differences among the many alternative proposals. Some of the proposed models have
been developed largely on the basis of theoretical reasoning informed by research (e.g., the TCI, MIPS, and

42

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

four clinical spectra), whereas others were developed


empirically through analyses of systematically sampled sets of personality traits or symptoms (e.g.,
DAPP-BQ, FFM, IPC, SNAP, SWAP-200, and PSY-5).
The models can also be differentiated with respect to
whether they are confined largely to personality disorder symptoms (e.g., DAPP-BQ, SNAP, and SWAP200); whether they include a full range of normal and
abnormal personality functioning (e.g., FFM, IPC,
TCI, and MIPS); and whether they also include Axis I
symptoms, e.g., the four clinical spectra and SNAP).
The models also differ with respect to their hierarchical level of description. Some of the models are confined to broad domains of personality functioning
(e.g., the four clinical spectra, the three MIPS polarities, the Zuckerman five dimensions, and the PSY-5),
whereas others include lower-order traits within a hierarchical structure (e.g., the DAPP-BQ, SNAP, FFM,
and TCI).

Common Higher-Order Domains


Fortunately, most of the alternative models do appear
to be readily integrated within a common hierarchical
structure (Bouchard and Loehlin 2001; John and
Srivastava 1999; Krueger and Tackett 2003; Larstone et
al. 2002; Livesley 2003; Zuckerman 2002). This common structure is hardly surprising, because most of
them are attempting to do largely the same thing (i.e.,
identify the fundamental dimensions of maladaptive
personality functioning that underlie and cut across
the existing diagnostic categories). Table 31 lists how
the broad domains of the DAPP-BQ, FFM, SNAP,
MPQ, PSY-5, IPC, Eysenck (1987), Zuckerman (2002),
Siever and Davis (1991), Tyrer (1988), and Cloninger
(2000) models might be aligned with one another. The
self-other, pleasure-pain, and active-passive polarity
model of Millon et al. (1996) is not included in the table
because its alignment with the other models is ambiguous and because only one study has empirically related these polarities to the other models (Millon
1994). The placement of Cloningers (2000) model is
also perhaps relatively more difficult than the others
(De Fruyt et al. 2000; Zuckerman 2002).
It is evident from Table 31 that all of the models
include a domain that concerns extraversion, otherwise described as sociability, activity, positive emotionality, or inhibition (when keyed in the negative direction). This domain contrasts being gregarious,
talkative, assertive, and active with being withdrawn,
isolated, introverted, and anhedonic. The terms extraversion and positive emotionality might appear to sug-

gest different domains of personality functioning.


However, many studies have confirmed that these are
in fact the same domains (Bouchard and Loehlin 2001;
Harkness et al. 1995; John and Srivastava 1999; Watson
et al. 1994). The title positive affectivity is preferred by
some authors because it is believed that positive affectivity might be providing the motivating force for extraversion, reflecting individual differences in a behavioral activation (or reward sensitivity) system
(Depue and Collins 1999; Pickering and Gray 1999;
Watson and Clark 1997). The Zuckerman domains of
sociability and activity and the Siever and Davis domain of inhibition are italicized in Table 31 because
they are relatively more narrow in their scope and coverage. Neither agency nor communion from the IPC
are aligned directly under this domain because they
are 45-rotated versions of extraversion and agreeableness (Wiggins 2003).
All of the dimensional models also include traits
referring to aggressive, dissocial, or antagonistic interpersonal relatedness. This domain contrasts being suspicious, rejecting, exploitative, antagonistic, callous,
deceptive, and manipulative with being trusting, compliant, agreeable, modest, dependent, diffident, and
empathic. This domain is represented more narrowly
by the PSY-5 and by Zuckerman because their versions
of this domain are confined largely to interpersonal
aggressiveness, whereas the other models include
such additional components as mistrust, exploitation,
suspiciousness, deception, and arrogance. Psychoticism from Eysencks dimensional model is not aligned
perfectly with this domain because he includes within
psychoticism both interpersonal antagonism and
impulsive disinhibition (Bouchard and Loehlin 2001;
Eysenck 1987; John and Srivastava 1999), comparable
with the conceptualization of this domain by Siever
and Davis. It should also be noted that the title psychoticism is perhaps somewhat unusual, because this term
is more typically understood to refer to cognitiveperceptual aberrations (as it is understood within the
PSY-5).
The three-dimensional models of the MPQ and
SNAP do not include an antagonistic, aggressive domain of personality functioning at this higher-order
level. The SNAP does include scales for mistrust, manipulativeness, and aggression but these are placed
within the domain of negative affectivity, and the
MPQ includes an aggression scale within the domain
of negative emotionality. However, joint factor analyses of the DAPP-BQ and SNAP subscales have yielded
consistently a four-factor solution (Clark and Livesley
2002; Clark et al. 1996) that corresponds to the first

Table 31.

Alignment of alternative dimensional models: broad domains


First

Second

Third

Fourth

Inhibition

Dissocial

Compulsivity

Emotional dysregulation

Five-factor model

Extraversion

Antagonism

Conscientiousness

Neuroticism

SNAP and MPQ

Positive affectivity

(Negative affectivity)

Constraint

Negative affectivity

PSY-5

Positive emotionality

Aggressiveness

Constraint

Negative emotionality

DAPP-BQ

Psychoticism

Agency
Communion

Eysenck

Extraversion

Zuckerman

Sociability
Activity

Aggression-Hostility

Tyrer

Withdrawn

Antisocial-Dependent

Siever and Davis

(Inhibition)

TCI

Openness

Psychoticism
Impulsive

Neuroticism
Inhibited

Aggression/Impulsivity
Cooperativeness

Neuroticism

Persistence

Reward dependence

Affective instability
Anxiety/Inhibition

Cognitive/Perceptual

Harm avoidance
Self-directedness

Self-transcendence

Novelty seeking
Note. Selected scales from the IPC, Eysenck, Siever and Davis, and Cloninger models are off-center because they lie between the domains defined by the adjoining columns. Selected scales from
the SNAP, PSY-5, Zuckerman, Siever and Davis, and TCI models are italicized because they describe domains that are somewhat narrower in scope. Selected scales from the SNAP, Siever and
Davis, and TCI models are noted parenthetically because they are more strongly related to another domain. Selected scales from the DAPP-BQ, Zuckerman, Tyrer, and Siever and Davis include
the symbol because they are keyed in the opposite direction of the other scales. DAPP-BQ=Dimensional Assessment of Personality PathologyBasic Questionnaire; IPC=interpersonal circumplex;
MPQ=Multidimensional Personality Questionnaire; PSY-5=Personality PsychopathologyFive; SNAP=Schedule for Nonadaptive and Adaptive Personality; TCI=Temperament and Character
Inventory.

Categorical and Dimensional Models of Personality Disorders

IPC

Fifth

43

44

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

four domains of Table 31. As indicated by Watson et


al. (1994), extensive data indicate thatthe Big Three
and Big Five models define a common Big Four
space (p. 24), consisting of negative affectivity (neuroticism), positive affectivity (extraversion), antagonism, and constraint.
All but two of the models also include a domain
concerned with the control and regulation of behavior,
referred to as constraint, compulsivity, and conscientiousness or, when keyed in the opposite direction, impulsivity and disinhibition. This domain contrasts being disciplined, compulsive, dutiful, conscientious,
deliberate, workaholic, and achievement-oriented
with being irresponsible, lax, impulsive, negligent,
and hedonistic. The only models not to include this
domain of personality functioning are the IPC and Tyrer s (1988) four-domain model. Tyrer placed the
symptoms of the obsessive-compulsive (anankastic)
personality disorder within his inhibited domain,
which is defined largely by traits of anxiousness and
dysphoria (i.e., a different meaning for the term inhibition than is used by the DAPP-BQ). The IPC does not
include constraint versus disinhibition because it is a
two-dimensional model confined to interpersonal relatedness.
Finally, it is also evident from Table 31 that all but
one of the models includes a broad domain of emotional dysregulation, otherwise described as negative
affectivity or neuroticism. The domain of emotional
dysregulation contrasts feeling anxious, depressed,
despondent, labile, helpless, self-conscious, and vulnerable (and within some models, feeling angry) with
feeling invulnerable, self-assured, and perhaps even
glib, shameless, and fearless. The only model not to include this domain of personality functioning is again
the IPC. This fourth domain is also somewhat more
narrowly defined by Siever and Davis (1991) because
they separate anxiousness from affective instability.
In summary, the predominant models of normal
and abnormal personality functioning do appear to
converge onto four broad domains of personality functioning that can be described as extraversion versus introversion, antagonism versus agreeableness, constraint versus impulsivity, and emotional dysregulation
versus emotional stability. The authors of these various
models would not all agree on the best names for each
dimension, due in part to the fact that no single name is
likely to optimally describe an entire domain. Some
models place more emphasis on the normal variants
(e.g., NEO-PI-R and TCI), whereas other models place
more emphasis on the abnormal variants (e.g., DAPPBQ and SNAP). Finally, the models vary in how broadly

or narrowly they define each domain. Nevertheless, the


convergence among them is quite evident with respect
to the existence of the four domains. Empirical support
for the convergence of these models within a four-factor
structure has been provided in a number of studies
(e.g., Austin and Deary 2000; Clark et al. 1996; Deary et
al. 1998; Livesley et al. 1998; Mulder and Joyce 1997),
and perhaps even within some of the earliest, original
efforts to develop dimensional models of personality
disorder by Presly and Walton (1973) and Tyrer and Alexander (1979).
Only three of the models include a fifth broad domain, characterized within the FFM as openness to experience (or as unconventionality), within the PSY-5 as
psychoticism (i.e., illusions, misperceptions, perceptual aberrations, and magical ideation), and by Siever
and Davis (1991) as cognitive-perceptual aberrations.
Subscales within the SNAP (e.g., schizotypal thought),
DAPP-BQ (perceptual cognitive distortion), and the
MPQ (absorption) relate empirically to FFM unconventionality (Bouchard and Loehlin 2001; Clark and
Livesley 2002). A domain of openness is obtained in
joint factor analytic studies that provide sufficient representation of the domain (e.g., Clark and Livesley
2002). However, it appears to be the case that when
this domain of openness or unconventionality is narrowly defined as simply cognitive-perceptual aberrations, scales to assess the domain either load on other
factors (typically negative affectivity) or they define a
factor that is so small that it might not appear to be
worth identifying (Austin and Deary 2000; Clark et al.
1996; Larstone et al. 2002). Openness to experience is
itself the fifth and smallest domain of the FFM (Goldberg 1993). It is also possible that cognitive-perceptual
aberrations do not belong within a dimensional model
of normal and abnormal personality functioning, consistent with the ICD-10 inclusion of schizotypal as a
variant of schizophrenia rather than a personality disorder.
Note that Table 31 does not include the proposals
of Oldham and Skodol (2000), Tyrer and Johnson
(1996), or Westen and Shedler (2000), because the
models provided in the table concern dimensions of
maladaptive (and at times also adaptive) personality
functioning that, for the most part, cut across the existing diagnostic categories. Some personality disorders
might be confined largely to one broad domain (e.g.,
schizoid within the introversion domain and obsessive-compulsive within the compulsivity domain),
but most are more aptly described in terms of more
than one domain (e.g., antisocial personality disorder
would be represented by antagonism and disinhibi-

Categorical and Dimensional Models of Personality Disorders

Table 32.

45

Lower-order facets and diagnostic criteria within the domain of antagonism versus agreeableness

Abnormal high traits


DAPP-BQ:
SNAP:

Suspiciousness, interpersonal disesteem, conduct problems, passive oppositionality, rejection,


narcissism
Mistrust, manipulativeness, aggression, entitlement

DSM-IV-TR diagnostic criteria


Antisocial:

Unlawful behaviors, lying, aliases, physical fights, lacks remorse, deceitfulness

Paranoid:

Recurrent suspicions, preoccupation with doubts about loyalty or trustworthiness,


reluctance to confide in others, reading hidden demeaning or threatening messages,
persistent bearing of grudges, perceptions of attacks on character that are not apparent
to others

Narcissistic:

Arrogant attitudes, sense of entitlement, interpersonally exploitative, preoccupation


with fantasies of unlimited success, grandiose sense of self-importance, lack of
empathy

Schizotypal:

Suspicious or paranoid ideation

Normal high traits


NEO-PI-R:

Skepticism, self-confidence, tough-mindedness, cunning, shrewd, competitive

Normal low traits


NEO-PI-R:
TCI:

Trust, straightforwardness, altruism, compliance, modesty, tender-mindedness, agreeableness


Helpfulness, compassion, pure-heartedness, sentimentality, empathy

Abnormal low traits


DAPP-BQ: Diffidence
SNAP:
Dependency
TCI:

Dependence
DSM-IV-TR diagnostic criteria
Dependent:
Histrionic:

Difficulty expressing disagreement, difficulty making everyday decisions without


excessive amount of advice
Suggestible, easily influenced by others

Note. DAPP-BQ =Dimensional Assessment of Personality PathologyBasic Questionnaire; NEO-PI-R= NEO Personality Inventory
Revised; SNAP=Schedule for Nonadaptive and Adaptive Personality; TCI=Temperament and Character Inventory.

tion, avoidant by neuroticism and introversion, and


dependent by agreeableness and neuroticism). The
representation of the DSM-IV-TR personality disorders becomes more evident when the lower-order facets of each domain are articulated.

Lower-Order Traits and Symptoms


Some of the dimensional models include lower-order
scales beneath the four (or five) broad domains of personality functioning. Table 32 provides a description
of how the respective personality trait scales from the
DAPP-BQ, SNAP, TCI, and FFM within the domain of
agreeableness versus antagonism are aligned with one
another, along with the respective personality disorder diagnostic criteria that correspond to these personality traits.
The alignment of the lower-order scales is helpful
in illustrating the hierarchical relationship among the

domains, traits, and behavioral diagnostic criteria. All


of the lower-order scales included within Table 32
(i.e., DAPP-BQ, SNAP, NEO-PI-R, and TCI scales) have
been shown empirically to be organized within a
higher-order domain of antagonism versus agreeableness (De Fruyt et al. 2000; Reynolds and Clark 2001),
but one can also proceed even lower in the hierarchy to
the level of the behavioral symptoms or expressions of
these traits, as illustrated by diagnostic criteria from
the antisocial, paranoid, narcissistic, schizotypal, dependent, and histrionic personality disorders. For example, it is evident that antisocial lying is a behavioral
example of the broader trait of manipulation, and reading hidden or demeaning messages in statements by
others is a more specific expression of the general trait
of mistrust or suspiciousness. Some DSM-IV-TR diagnostic criteria, however, are also at the level of the personality traits (e.g., sense of entitlement) rather than
being specific behavioral acts (Clark 1992; Shea 1992).

46

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Table 32 is also useful in illustrating the close relationship of the normal and abnormal variants of
these traits. Scales from the NEO-PI-R and TCI refer
largely to normal variants of agreeableness (i.e., being
trusting, compliant, straightforward, altruistic, modest, helpful, compassionate, sentimental, and empathic), whereas the scales from the DAPP-BQ and
SNAP refer largely to abnormal, maladaptive variants of these same traits (i.e., being dependent, diffident, gullible, sacrificial, meek, docile, submissive, or
self-denigrating).
Finally, Table 32 also illustrates normal and abnormal variants at both of the antagonism and agreeableness poles of this domain of personality functioning. There are abnormal variants of being excessively
high in antagonism (e.g., suspicious, aggressive, or
callous) and abnormal variants for the opposite pole,
being excessively high in agreeableness (e.g., diffidence, dependency, gullibility, and meekness). There
are maladaptive variants for both poles of all of the domains of general personality functioning (Coker et al.
2002; Trull et al. 1998).
Table 33 provides a comparable illustration for the
domain of emotional dysregulation versus emotional
stability. In this instance, the lack of a clear boundary
between the normal and abnormal variants is even
more apparent, particularly for the low levels of emotional dysregulation. For example, the neuroticism
scales of the NEO-PI-R assess levels of anxiousness, depressiveness, self-consciousness, and vulnerability
that are present within persons of the general population who would not typically be diagnosed as having a
personality disorder, whereas the anxiousness scale
from the DAPP-BQ was derived from studies of maladaptive personality functioning. The maladaptivity
of the most extreme expressions of normal anxiousness, depressiveness, helplessness, and vulnerability,
however, are self-evident, as in the suicidal behavior
and self-mutilation evident within persons diagnosed
with BPD.
Table 33 also illustrates that one can even identify
maladaptive variants of extremely high emotional
regulation (i.e., the lower half of Table 33), evident in
psychopathic persons who may lack the ability to experience normal adaptive feelings of vulnerability,
anxiousness, or self-consciousness (Hare 1991; Hare et
al. 1991). Cleckley (1976) had included in his original
description of psychopathy an absence of nervousness or psychoneurotic manifestations (p. 206). The
psychopath is nearly always free from minor reactions
popularly regarded as neurotic or as constituting
nervousness (p. 54), contributing perhaps to the

(unself-conscious) glib charm of the psychopath and


to the failure to adequately experience signs of threat
or to respond effectively to punishment, and to feelings of invulnerability and invincibility (Lykken 1995;
Lynam 2002).
Table 34 provides trait terms and diagnostic
criteria from the domain of constraint versus disinhibition. Normal and abnormal variants of constraint
are again readily identified, with a number of scales
from the TCI and the NEO-PI-R that refer to normal,
adaptive levels of constraint (or conscientiousness)
such as dutifulness, responsibility, ambitiousness,
resourcefulness, deliberation, and self-discipline
with maladaptive variants of these traits emphasized
by the DAPP-BQ and the SNAP (i.e., compulsivity,
workaholism, and propriety) that are in turn evident
within the more behavioral diagnostic criteria for the
obsessive-compulsive personality disorder (e.g., excessive devotion to work and preoccupation with details, rules, and organization). At the opposite pole of
the constraint domain are the impulsivity and disorderliness scales from the SNAP and TCI and the disinhibited, lax, negligent, disorderly, and irresponsible
behaviors of the antisocial and passive-aggressive personality disorders.

CLINICAL UTILITY
Categorical models of classification are often preferred because they appear to be easier to use (Frances
et al. 1995). One diagnostic label can convey a considerable amount of useful information in a vivid and
succinct manner. Dimensional models of classification
are, in one respect, inherently more complex than diagnostic categories because they generally provide
more specific and precise information. For example, it
is simpler to inform a colleague that a patient has BPD
than to describe the patient in terms of the 30 facets of
the FFM.
However, the existing diagnostic categories are
frustrating and troublesome to clinicians in part because the simplicity of the categorical model provides
inaccurate and misleading descriptions (Kass et al.
1985; Maser et al. 1991). Clinicians could find a dimensional model of classification to be easier to use because it provides a more valid and internally consistent means with which to describe a particular
patients psychopathology (Kass et al. 1985). A dimensional classification could be less cumbersome because it would not require the assessment of numerous diagnostic criteria from overlapping categories in

Categorical and Dimensional Models of Personality Disorders

Table 33.

47

Lower-order traits, facets, and diagnostic criteria within the domain of emotional dysregulation versus
emotional stability

Abnormal high traits


DAPP-BQ:

Affective lability, self-harm, anxiousness, identity problems, (insecure attachment), (intimacy


problems), (social avoidance)

SNAP:

Suicide potential, (dependency)

DSM-IV diagnostic criteria


Borderline:

Affective instability, recurring suicidal behavior, unstable and intense relationships, frantic
efforts to avoid abandonment, inappropriate and intense anger

Avoidant:

Fear of being shamed or ridiculed, feelings of inadequacy, view of self as inept or inadequate

Dependent:

Preoccupation with fears of being alone, losing support, or being left to care of self

Schizotypal:

Social anxiety

Normal high traits


NEO-PI-R:

Self-consciousness, anxiousness, depressiveness, vulnerability, responsive

TCI:

Shyness, worry/pessimism, fear of uncertainty

Normal low traits


NEO-PI-R:

Calm, low self-consciousness, self-assured, relaxed, resilient

TCI:

Self-acceptance

Abnormal low traits


DAPP-BQ:

(Narcissism)

PCL-R:

Glib and superficial charm

Personality disorder diagnostic criteria


Psychopathic: Shamelessness, fearlessness, feelings of invulnerability or invincibility, inability to feel anxious
Note. Some scales from the DAPP-BQ and SNAP are noted parenthetically because they include aspects of personality function from another domain. DAPP-BQ=Dimensional Assessment of Personality PathologyBasic Questionnaire; NEO-PI-R=NEO Personality Inventory
Revised; PCL-R=Hare Psychopathy ChecklistRevised; SNAP=Schedule for Nonadaptive and Adaptive Personality; TCI=Temperament
and Character Inventory.

a frustratingly unsuccessful effort to make illusory


distinctions. Semistructured interviews for the DSMIV-TR personality disorders must evaluate approximately 100 diagnostic criteria, whereas a semistructured interview for the FFM that covers both normal
and maladaptive personality functioning requires the
assessment of only 30 facets of personality functioning
(Trull et al. 1998). A dimensional model of classification would have an immediate benefit to clinical practice through its resolution of the problems of diagnostic co-occurrence, heterogeneity of membership,
inconsistent and ill-defined diagnostic boundaries, inadequate coverage, and illusory diagnostic distinctions.
A potential limitation of some of the dimensional
models is the absence of much literature on the treatment implications for elevations on some of the respective scales, or at least an absence of familiarity
among clinicians with respect to this literature (Sprock
2003). For example, many clinicians might feel lost
when informed that their client has maladaptively low

or high levels of TCI persistence, FFM altruism, or


MIPS active instrumental behavior. On the other hand,
the dimensional models of personality disorder that
are closest to the existing diagnostic categories (e.g.,
DAPP-BQ and SNAP) are readily able to draw upon
the extensive clinical literature concerning the treatment implications of each personality disorder. Very
little additional training would be necessary for the
clinician to apply the profile descriptions proposal of
Oldham and Skodol (2000), Tyrer and Johnson (1996),
or Westen and Shedler (2000).
In addition, it is also apparent from Tables 32
through 34 that a dimensional model could retain the
existing personality disorder symptoms as lower-order
(behavioral) manifestations of a respective personality
trait. Clinicians familiar with the treatment of borderline suicidal behavior, avoidant social anxiety, dependent feelings of inadequacy, or paranoid recurrent suspiciousness would still be treating these symptoms,
and a dimensional model of personality disorder could
still refer explicitly to them. The major difference

48

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Table 34.

Lower-order traits, facets, and diagnostic criteria within the domain of constraint versus disinhibition

Abnormal high traits


DAPP-BQ:

Compulsivity

SNAP:

Workaholism, propriety

TCI:

Perfectionism, work-hardened
DSM-IV diagnostic criteria
Obsessive-compulsive: Preoccupation with details, rules, lists, order and organization; perfectionism;
excessive devotion to work; overly conscientious; scrupulous; unable to discard
worn-out or worthless objects

Normal high traits


NEO-PI-R:

Dutifulness, order, achievement striving, self-discipline, deliberation, competence

TCI:

Resourcefulness, eagerness of effort, responsibility, ambitiousness, purposefulness

Normal low traits


NEO-PI-R:

Casual, easygoing, intuitive, playful

Abnormal low traits


DAPP-BQ:

(Conduct problems), (stimulus-seeking), (passive-oppositionality)

SNAP:

Impulsivity

TCI:

Disorderliness
DSM-IV diagnostic criteria
Passive-aggressive:

Passive resistance to fulfilling routine social and occupational tasks

Antisocial:

Impulsivity, failure to plan ahead, consistent irresponsibility, recklessness

Note. Some scales from the DAPP-BQ are noted parenthetically because they include aspects of personality function from another domain.
DAPP-BQ= Dimensional Assessment of Personality PathologyBasic Questionnaire; NEO-PI-R = NEO Personality InventoryRevised;
SNAP=Schedule for Nonadaptive and Adaptive Personality; TCI=Temperament and Character Inventory.

would just be that the dimensional models would provide these symptoms within dimensions that would be
appreciably less overlapping than the existing diagnostic categories.
The personality domain organization provided in
Table 31 could in fact facilitate treatment recommendations, as each domain would have more differentiated implications for functioning and treatment
planning than the existing diagnostic categories. For
example, the first two domains concern disorders of
interpersonal relatedness that would be of particular
interest and concern to clinicians specializing in marital or family therapy. The third domain involves, at
one pole, disorders of impulse dysregulation and disinhibition for which there is again a considerable
amount of treatment literature (Coccaro 1998). Disorders within this realm would be particularly evident
in behavior that affects work, career, and parenting,
with laxness, irresponsibility, and negligence at one
pole and a maladaptively excessive perfectionism and
workaholism at the other pole. The fourth domain
would be most suggestive of pharmacotherapy (as
well as psychotherapeutic) interventions for the treat-

ment of various forms of affective dysregulation that


are currently spread across the diagnostic categories,
including anxiousness, depressiveness, anger, and instability of mood. If the fifth domain of unconventionality was included, it would have specific implications
for impaired reality testing, magical thinking, and perceptual aberrations at one pole (Siever and Davis
1991) and perhaps alexithymia, closed-mindedness,
and a sterile absence of imagination at the other.
A dimensional model of classification would also
have the potentially useful advantage of providing
both adaptive and maladaptive personality traits. One
can indicate whether a patient is trusting, gregarious,
agreeable, and achievement striving, as well as whether
the patient shows the maladaptive variants of these
traits (i.e., gullibility, intolerance of being alone, docile
acquiescence, and workaholism, respectively). Clinicians can then not only describe their patients in a more
accurate and specific manner by indicating their precise
location along the various dimensions of personality
functioning, but also provide a more thorough and
comprehensive description by including traits that contribute to adaptive functioning and treatment respon-

Categorical and Dimensional Models of Personality Disorders

sivity. This comprehensive profile description would


then draw not only on the existing clinical literature
concerning affective dysregulation, impulsivity, workaholism, and interpersonal relatedness but also on the
basic science literature concerning the etiology and development of general personality functioning.

Case Example
This case is a summary of a woman, Ms. B, who participated in a dialectical behavior therapy (DBT) program, described by Sanderson and Clarkin (2002).
Ms. B was a 37-year-old, married Hispanic woman
with three children. She had a bachelors degree in
nursing but had been on psychiatric disability leave
for the past 2 years.
Ms. B had done well in school as a child, although she was at times a problem for her teachers
because she would occasionally seem to explode in
an inexplicable anger and tirade. She was the second
of eight children in a family in which there was quite
severe corporal punishment. Whenever her parents
discovered that she had been reprimanded at school,
she would be severely punished at home, at times
reaching the level of bruises, wounds, and scars. At
the age of 14, she began to be repeatedly sexually
abused by a friend of the family. The abuse ended
when it became known to her parents, but Ms. B felt
that they also considered her to be at least partly responsible. Her mother often prayed for her lost soul,
and her father often referred to her as the lost one.
As the second oldest child, she had considerable
household responsibilities, and she would often be
punished severely for failing to meet them. She described having very mixed feelings toward her
mother, feeling that she let her mother down yet also
feeling bitter and angry in not being adequately protected from the sexual abuser or her physically abusive father.
Ms. B had been hospitalized seven times prior to
her entry into the DBT program. Her previous diagnoses included major depressive disorder, posttraumatic stress disorder, generalized anxiety disorder,
and BPD. She was given a diagnosis of BPD upon entry into the DBT program, but it did not appear to
her therapist that this diagnosis adequately described her difficulties or her strengths. From the
perspective of the dimensional model of description
of Table 31, she clearly had difficulty with affective
regulation. She would be expected to have elevations on the DAPP-BQ scales for affective lability
and self-harm, the SNAP scale for suicide potential,
and perhaps the DAPP-BQ scales for identity problems and insecure attachment as well (see Table 33).
She completed the self-report NEO-PI-R inventory
(Costa and McCrae 1992), which indicated substantial elevations on anxiousness, depressiveness, angry hostility, and vulnerability. She also obtained
markedly low elevations on facets of agreeableness
(compliance and straightforwardness) that are commonly seen in persons diagnosed with BPD (Clarkin

49

et al. 1993), indicating defiance and manipulative deception. However, inconsistent with these expressions of antagonism were adaptive elevations on the
agreeableness facets of modesty and altruism and
the extraversion facet of warmth. Ms. B was often
defiant, oppositional, and angry, particularly toward
people in authority, but she was also very self-sacrificial, self-denying, and self-deprecating (Sanderson and Clarkin 2002, p. 367). The borderline diagnosis did not do justice to these specific aspects of her
personality. She would often get into verbal fights
and arguments, but these arguments were also coupled with sincere feelings of warmth and concern toward others (p. 367).
Of particular importance to her entry into the
DBT program were her adaptive elevations within
the domain of constraint (conscientiousness; see Table 34). Elevations on facets of conscientiousness
are not usually seen in patients with BPD, but they
bode well for a potential responsivity to the rigors
and demands of the DBT program (Sanderson and
Clarkin 2002, p. 367). On the one hand, Ms. B had a
very dysfunctional life, with seven hospitalizations
and loss of employment due to a psychiatric disability. On the other hand, she had also accomplished a
great deal despite her abusive past and negative
emotionality, including good grades in school, a
bachelors degree, a (temporarily suspended) nursing career, and a successful marriage. She clearly
did aspire to be successful and competent in all that
she did (p. 367), including the DBT program. Ms. B
responded well to the DBT social skills group and
eventually even became a mentor to the younger patients within the group.
Complicating her involvement in the DBT program, however, were her relatively low scores on the
NEO-PI-R scales for openness to values and ideas.
Ms. B came from a relatively conservative background, and she had an unwavering attitude regarding many matters of life. Fundamental to her depressiveness was her self-deprecation and self-blame,
but she was also highly resistant to questioning this
self-criticism. Her therapist eventually abandoned
her effort to confront Ms. Bs strong moral attitudes,
focusing instead on developing a forgiveness of others for the pain she had suffered at their hands.

CONCLUSIONS AND RECOMMENDATIONS


The description and classification of personality disorders currently use a categorical model, wherein a person
is provided with a single diagnostic label to describe his
or her maladaptive personality traits. However, it appears that personality disorders, like general personality functioning, are not summarized well by one single
diagnostic label. Persons appear instead to have constellations of maladaptive (and adaptive) personality traits
that might be better described in terms of dimensional

50

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

models. A number of alternative dimensional models of


personality disorder have now been developed, and it
appears that most of them can be readily integrated into
a common hierarchical structure.

REFERENCES
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition. Washington,
DC, American Psychiatric Association, 1980
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition, Revised.
Washington, DC, American Psychiatric Association,
1987
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association, 2000
Ashton MC, Lee K, Goldberg LR: A hierarchical analysis of
1,710 English personality-descriptive adjectives. J Pers
Soc Psychol 87:707721, 2004
Austin EJ, Deary IJ: The four As: a common framework for
normal and abnormal personality? Pers Individ Dif
28:977995, 2000
Basic Behavioral Science Task Force of the National Advisory Mental Health Council: Basic behavioral science
research for mental health: vulnerability and resilience.
Am Psychol 51:2228, 1996
Benjamin LS: Interpersonal Diagnosis and Treatment of Personality Disorders, 2nd Edition. New York, Guilford,
1996
Blashfield RK, Blum N, Pfohl B: The effects of changing Axis
II diagnostic criteria. Compr Psychiatry 33:245252,
1992
Bornstein RF: Reconceptualizing personality disorder diagnosis in the DSM-V: the discriminant validity challenge.
Clin Psychol 5:333343, 1998
Bouchard TJ, Loehlin JC: Genes, evolution, and personality.
Behav Genet 31:243273, 2001
Brinkley CA, Newman JP, Widiger TA, et al: Two approaches
to parsing the heterogeneity of psychopathy. Clin Psychol 11:6994, 2004
Cannon T, Turkheimer E, Oltmanns TF: Factorial structure of
pathological personality as evaluated by peers. J Abnorm Psychol 112:8191, 2003
Clark LA: Resolving taxonomic issues in personality disorders. J Personal Disord 6:360378, 1992
Clark LA: Manual for the Schedule for Nonadaptive and
Adaptive Personality. Minneapolis, MN, University of
Minnesota Press, 1993
Clark LA, Livesley WJ: Two approaches to identifying the
dimensions of personality disorder: convergence on the
five-factor model, in Personality Disorders and the
Five-Factor Model of Personality, 2nd Edition. Edited
by Costa PT, Widiger TA. Washington, DC, American
Psychological Association, 2002, pp 161176

Clark LA, Watson D: Temperament: a new paradigm for trait


psychology, in Handbook of Personality: Theory and
Research, 2nd Edition. Edited by Pervin L, John O. New
York, Guilford, 1999, pp 399423
Clark LA, Livesley WJ, Schroeder ML, et al: Convergence of
two systems for assessing personality disorder. Psychol
Assess 8:294303, 1996
Clarkin JF, Hull JW, Cantor J, et al: Borderline personality
disorder and personality traits: a comparison of SCID-II
BPD and NEO-PI. Psychol Assess 5:472476, 1993
Cleckley H: The Mask of Sanity, 5th Edition. St Louis, MO,
Mosby, 1976
Cloninger CR: The genetics and psychobiology of the sevenfactor model of personality, in Biology of Personality
Disorders. Edited by Silk KR. Washington, DC, American Psychiatric Press, 1998, pp 6392
Cloninger CR: A practical way to diagnosis personality disorders: a proposal. J Personal Disord 14:99108, 2000
Cloninger CR, Svrakic D: Personality disorder, in Comprehensive Textbook of Psychiatry, 7th Edition. Edited by
Sadock B, Kaplan H. Baltimore, MD, Williams and Wilkins, 1999, pp 15671588
Coccaro EF: Neurotransmitter function in personality disorders, in Biology of Personality Disorders. Edited by Silk
KR. Washington, DC, American Psychiatric Press, 1998,
pp 125
Coker LA, Samuel DB, Widiger TA: Maladaptive personality
functioning within the Big Five and the FFM. J Personal
Disord 16:385401, 2002
Costa PT, McCrae RR: Revised NEO Personality Inventory
(NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI)
Professional Manual. Odessa, FL, Psychological Assessment Resources, 1992
Deary IJ, Peter A, Austin E, et al: Personality traits and personality disorders. Br J Psychol 89:647661, 1998
De Fruyt F, van De Wiele L, van Heeringen C: Cloningers
psychobiological model of temperament and character
and the five-factor model of personality. Pers Individ
Dif 29:441452, 2000
Depue RA, Collins PF: Neurobiology of the structure of personality: dopamine facilitation of incentive motivation
and extraversion. Behav Brain Sci 22:491569, 1999
Eysenck HJ: The definition of personality disorders and the
criteria appropriate for their description. J Personal Disord 1:211219, 1987
First MB, Bell CB, Cuthbert B, et al: Personality disorders and
relational disorders: a research agenda for addressing
crucial gaps in DSM, in A Research Agenda for DSM-V.
Edited by Kupfer DJ, First MB, Regier DA. Washington,
DC, American Psychiatric Association, 2002, pp 123199
Frances AJ, First MB, Pincus HA: DSM-IV Guidebook.
Washington, DC, American Psychiatric Press, 1995
Friedman HS, Tucker JS, Schwartz JE, et al: Childhood conscientiousness and longevity: health behaviors and
cause of death. J Pers Soc Psychol 68:696703, 1995
Goldberg LR: The structure of phenotypic personality traits.
Am Psychol 48:2634, 1993

Categorical and Dimensional Models of Personality Disorders

Gunderson JG: Diagnostic controversies, in American Psychiatric Press Review of Psychiatry, Vol 11. Edited by
Tasman A, Riba MB. Washington, DC, American Psychiatric Press, 1992, pp 924
Hare RD: The Hare Psychopathy ChecklistRevised Manual. North Tonawanda, NY, Multi-Health Systems, 1991
Hare RD, Hart SD, Harpur TJ: Psychopathy and the DSM-IV
criteria for antisocial personality disorder. J Abnorm
Psychol 100:391398, 1991
Harkness AR, McNulty JL, Ben Porath YS: The Personality
Psychopathology Five (PSY-5): constructs and MMPI-2
scales. Psychol Assess 7:104114, 1995
Haslam N: The dimensional view of personality disorders: a
review of the taxometric evidence. Clin Psychol Rev
23:7593, 2003
Herkov MJ, Blashfield RK: Clinicians diagnoses of personality disorder: evidence of a hierarchical structure.
J Pers Assess 65:313321, 1995
John OP, Srivastava S: The big five trait taxonomy: history,
measurement, and theoretical perspectives, in Handbook of Personality: Theory and Research, 2nd Edition.
Edited by Pervin LA, John OP. New York, Guilford, 1999,
pp 102138
Judge TA, Higgins CA, Thoresen CJ, et al: The big five personality traits, general mental ability, and career success
across the life span. Personnel Psychology 52:621652,
1999
Kass F, Skodol A, Charles E, et al: Scaled ratings of DSM-III
personality disorders. Am J Psychiatry 142:627630, 1985
Kiesler DJ: Contemporary Interpersonal Theory and Research: Personality, Psychopathology, and Psychotherapy. New York, Wiley, 1996
Klein DN, Riso LP: Psychiatric disorders: problems of boundaries and comorbidity, in Basic Issues in Psychopathology. Edited by Costello CG. New York, Guilford, 1993,
pp 1966
Klein MH, Benjamin LS, Rosenfeld R, et al: The Wisconsin
Personality Disorders Inventory, I: development, reliability, and validity. J Personal Disord 7:285303, 1993
Krueger RF: The structure of common mental disorders.
Arch Gen Psychiatry 56:921926, 1999
Krueger RF, Tackett JL: Personality and psychopathology:
working toward the bigger picture. J Personal Disord
17:109128, 2003
Larstone RM, Jang KL, Livesley WJ, et al: The relationship
between Eysencks P-E-N model of personality, the five
factor model of personality, and traits delineating personality dysfunction. Pers Individ Dif 33:2537, 2002
Lilienfeld SO, Waldman ID, Israel AC: A critical examination of the use of the term comorbidity in psychopathology research. Clin Psychol 1:7183, 1994
Livesley WJ: Diagnostic dilemmas in classifying personality
disorder, in Advancing DSM: Dilemmas in Psychiatric
Diagnosis. Edited by Phillips KA, First MB, Pincus HA.
Washington, DC, American Psychiatric Association,
2003, pp 153190

51

Livesley WJ, Jang KL, Vernon PA: Phenotypic and genetic


structure of traits delineating personality disorder. Arch
Gen Psychiatry 55:941948, 1998
Lykken DT: The Antisocial Personalities. Hillsdale, NJ,
Erlbaum, 1995
Lynam DR: Psychopathy from the perspective of the five factor model, in Personality Disorders and the Five Factor
Model, 2nd Edition. Edited by Costa PT, Widiger TA.
Washington DC, American Psychological Association,
2002, pp 325348
Lynam DR, Widiger TA: Using the five-factor model to represent the DSM-IV personality disorders: an expert consensus approach. J Abnorm Psychol 110:401412, 2001
Maser JD, Kaelber C, Weise RF: International use and attitudes toward DSM-III and DSM-III-R: growing consensus in psychiatric classification. J Abnorm Psychol 100:
271279, 1991
McCrae RR, Costa PT: A five-factor theory of personality, in
Handbook of Personality: Theory and Research, 2nd
Edition. Edited by Pervin LA, John OP. New York, Guilford, 1999, pp 139153
Miller JD, Lynam DR: Psychopathy and the five-factor
model of personality: a replication. J Pers Assess 81:168
178, 2003
Miller JD, Lynam DR, Widiger TA, et al: Personality disorders as extreme variants of common personality dimensions: can the five-factor model adequately represent
psychopathy? J Pers 69:253276, 2001
Millon T: Millon Index of Personality Styles Manual. San Antonio, TX, The Psychological Corporation, 1994
Millon T, Davis RD, Millon CM, et al: Disorders of Personality: DSM-IV and Beyond. New York, Wiley, 1996
Morey LC, Zanarini MC: Borderline personality: traits and
disorder. J Abnorm Psychol 109:733737, 2000
Mulder RT, Joyce PR: Temperament and the structure of personality disorder symptoms. Psychol Med 27:99106,
1997
Narrow WE, Rae DS, Robins LN, et al: Revised prevalence
estimates of mental disorders in the United States: using
a clinical significance criterion to reconcile 2 surveys estimates. Arch Gen Psychiatry 59:115123, 2002
OConnor BP: The search for dimensional structure differences between normality and abnormality: a statistical
review of published data on personality and psychopathology. J Pers Soc Psychol 83:962982, 2002
Oldham JM, Skodol AE: Charting the future of Axis II. J Personal Disord 14:1729, 2000
Oldham JM, Skodol AE, Kellman HD, et al: Diagnosis of
DSM-III-R personality disorders by two semistructured
interviews: patterns of comorbidity. Am J Psychiatry
149:213220, 1992
Pickering AD, Gray JA: The neuroscience of personality, in
Handbook of Personality: Theory and Research, 2nd
Edition. Edited by Pervin LA, John OP. New York, Guilford, 1999, pp 277299
Pincus AL, Wilson KR: Interpersonal variability in dependent personality. J Pers 69:223251, 2001

52

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Pincus HA, McQueen LE, Elinson L: Subthreshold mental


disorders: nosological and research recommendations,
in Advancing DSM: Dilemmas in Psychiatric Diagnosis.
Edited by Phillips KA, First MB, Pincus HA. Washington, DC, American Psychiatric Association, 2003, pp
129144
Plomin R, Caspi A: Behavioral genetics and personality, in
Handbook of Personality: Theory and Research, 2nd Edition. Edited by Pervin LA, John OP. New York, Guilford,
1999, pp 251276
Presly AS, Walton HJ: Dimensions of abnormal personality.
Br J Psychiatry 122:269276, 1973
Reynolds SK, Clark LA: Predicting dimensions of personality disorder from domains and facets of the Five-Factor
Model. J Pers 69:199222, 2001
Rounsaville BJ, Alarcn RD, Andrews G, et al: Basic nomenclature issues for DSM-V, in A Research Agenda for
DSM-V. Edited by Kupfer DJ, First MB, Regier DE. Washington, DC, American Psychiatric Association, 2002, pp
129
Samuel DB, Widiger TA: Clinicians personality descriptions
of prototypic personality disorders. J Personal Disord
18:286308, 2004
Sanderson C, Clarkin JF: Further use of the NEO PI-R personality dimensions in differential treatment planning, in
Personality Disorders and the Five Factor Model of Personality, 2nd Edition. Edited by Costa PT, Widiger TA.
Washington, DC, American Psychological Association,
2002, pp 351375
Sanislow CA, Morey LC, Grilo CM, et al: Confirmatory factor
analysis of DSM-IV borderline, schizotypal, avoidant
and obsessive-compulsive personality disorders: findings from the Collaborative Longitudinal Personality
Disorders Study. Acta Psychiatr Scand 105:2836, 2002
Saulsman LM, Page AC: The five-factor model and personality disorder empirical literature: a meta-analytic review. Clin Psychol Rev 23:10551085, 2004
Shea MT: Some characteristics of the Axis II criteria sets and
their implications for assessment of personality disorders. J Personal Disord 6:377381, 1992
Shedler J: A new language for psychoanalytic diagnosis.
J Am Psychoanal Assoc 50:429456, 2002
Shedler J, Westen D: Refining DSM-IV personality disorder
diagnosis: integrating science and practice. Am J Psychiatry 161:13501365, 2004
Siever LJ, Davis KL: A psychobiological perspective on the
personality disorders. Am J Psychiatry 148:16471658,
1991
Spitzer R, Endicott J, Gibbon M: Crossing the border into
borderline personality and borderline schizophrenia.
Arch Gen Psychiatry 36:1724, 1979
Sprock J: Dimensional versus categorical classification of
prototypic and nonprototypic cases of personality disorder. J Clin Psychol 59:9911014, 2003

Svrakic DM, Draganic S, Hill K, et al: Temperament, character, and personality disorders: etiologic, diagnostic, and
treatment issues. Acta Psychiatr Scand 106:189195,
2002
Tellegen A: Multidimensional Personality Questionnaire:
Manual for Administration, Scoring, and Interpretation. Minneapolis, MN, University of Minnesota Press
(in press)
Trull TJ, Widiger TA, Useda JD, et al: A structured interview
for the assessment of the five-factor model of personality. Psychol Assess 10:229240, 1998
Trull TJ, Widiger TA, Lynam DR, et al: Borderline personality disorder from the perspective of general personality
functioning. J Abnorm Psychol 112:193202, 2003
Tyrer P: Personality Disorders: Diagnosis, Management, and
Course. London, Wright, Butterworth and Co, 1988
Tyrer P, Alexander J: Classification of personality disorder.
Br J Psychiatry 135:163167, 1979
Tyrer P, Johnson T: Establishing the severity of personality
disorder. Am J Psychiatry 153:15931597, 1996
Verheul R, Widiger TA: A meta-analysis of the prevalence
and usage of the personality disorder not otherwise
specified (PDNOS) diagnosis. J Personal Disord 18:309
319, 2004
Watson D, Clark LA: Extraversion and its positive emotional
core, in Handbook of Personality Psychology. Edited by
Hogan R, Johnson J, Briggs S. New York, Academic
Press, 1997, pp 767793
Watson D, Clark LA, Harkness AR: Structures of personality
and their relevance to psychopathology. J Abnorm Psychol 103:1831, 1994
Watson D, Wiese D, Vaidya J, et al: The two general activation systems of affect: structural findings, evolutionary
considerations, and psychobiological evidence. J Pers
Soc Psychol 76:820838, 1999
Westen D, Arkowitz-Westen L: Limitations of Axis II in diagnosing personality pathology in clinical practice. Am
J Psychiatry 155:17671771, 1998
Westen D, Shedler J: A prototype matching approach to diagnosing personality disorders: toward DSM-V. J Personal Disord 14:109126, 2000
Widiger TA: Four out of five aint bad. Arch Gen Psychiatry
55:865866, 1998
Widiger TA, Corbitt E: Normal versus abnormal personality
from the perspective of the DSM, in Differentiating Normal and Abnormal Personality. Edited by Strack S, Lorr
M. New York, Springer, 1994, pp 158175
Widiger TA, Costa PT: Five factor model personality disorder research, in Personality Disorders and the Five Factor Model of Personality, 2nd Edition. Edited by Costa
PT, Widiger TA. Washington, DC, American Psychological Association, 2002, pp 5987
Widiger TA, Hagemoser S: Personality disorders and the interpersonal circumplex, in Circumplex Models of Personality and Emotions. Edited by Plutchik R, Conte HR.
Washington, DC, American Psychological Association,
1997, pp 299325

Categorical and Dimensional Models of Personality Disorders

Widiger TA, Sanderson CJ: Towards a dimensional model of


personality disorders in DSM-IV and DSM-V, in The
DSM-IV Personality Disorders. Edited by Livesley WJ.
New York, Guilford, 1995, pp 433458
Widiger TA, Trull TJ: Performance characteristics of the DSMIII-R personality disorder criteria sets, in DSM-IV Sourcebook, Vol 4. Edited by Widiger TA, Frances AJ, Pincus
HA, et al. Washington, DC, American Psychiatric Association, 1998, pp 357373
Widiger TA, Costa PT, McCrae RR: A proposal for Axis II: diagnosing personality disorders using the five factor
model, in Personality Disorders and the Five Factor
Model of Personality, 2nd Edition. Edited by Costa PT,
Widiger TA. Washington, DC, American Psychological
Association, 2002, pp 431456

53

Wiggins JS: Paradigms of Personality Assessment. New


York, Guilford, 2003
World Health Organization: The ICD-10 Classification of
Mental and Behavioural Disorders: Clinical Descriptions
and Diagnostic Guidelines. Geneva, Switzerland, World
Health Organization, 1992
Zimmerman M, Mattia JI: Psychiatric diagnosis in clinical
practice: is comorbidity being missed? Compr Psychiatry 40:182191, 1999
Zuckerman M: Zuckerman-Kuhlman Personality Questionnaire (ZKPQ): an alternative five-factorial model, in Big
Five Assessment. Edited by de Raad B, Perugini M. Kirkland, WA, Hogrefe and Huber, 2002, pp 377397

This page intentionally left blank

Part II
Clinical Evaluation

This page intentionally left blank

4
Manifestations, Clinical
Diagnosis, and Comorbidity
Andrew E. Skodol, M.D.

A personality disorder is defined in DSM-IV-TR as an


enduring pattern of inner experience and behavior
that deviates markedly from the expectations of the individuals culture, is pervasive and inflexible, has an
onset in adolescence or early adulthood, is stable over
time, and leads to distress or impairment (American
Psychiatric Association 2000, p. 685). Personality disorders are reported on Axis II of the DSM-IV-TR multiaxial system to ensure that consideration is given to their
presence in all patient evaluations, even when Axis I
disorder psychopathology is present and prominent.
DSM-IV-TR includes criteria for the diagnosis of 10
specific personality disorders, arranged into three
clusters based on descriptive similarities. Cluster A is
commonly referred to as the odd or eccentric cluster
and includes paranoid, schizoid, and schizotypal personality disorders. Cluster B, the dramatic, emotional, or erratic cluster, includes antisocial, borderline, histrionic, and narcissistic personality disorders.
Cluster C, the anxious and fearful cluster, includes

avoidant, dependent, and obsessive-compulsive personality disorders. DSM-IV-TR also provides for a residual category of personality disorder not otherwise
specified (PDNOS). This category is to be used when a
patient meets the general criteria for a personality disorder and has features of several different types but
does not meet criteria for any specific personality disorder (i.e., mixed personality disorder) or is considered to have a personality disorder not included in the
official classification (e.g., self-defeating or depressive
personality disorders).

DEFINING FEATURES OF
PERSONALITY DISORDERS
Patterns of Inner Experience and Behavior
The general diagnostic criteria for a personality disorder in DSM-IV-TR (see Table 41) indicate that a pattern

Sections of this chapter have been modified with permission from Skodol AE: Problems in Differential Diagnosis: From DSM-III
to DSM-III-R in Clinical Practice. Washington, DC, American Psychiatric Press, 1989

57

58

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Table 41.
A.

General diagnostic criteria for a personality disorder

An enduring pattern of inner experience and behavior that deviates markedly from the expectations of the individual's
culture. This pattern is manifested in two (or more) of the following areas:
(1)
(2)
(3)
(4)

cognition (i.e., ways of perceiving and interpreting self, other people, and events)
affectivity (i.e., the range, intensity, lability, and appropriateness of emotional response)
interpersonal functioning
impulse control

B. The enduring pattern is inflexible and pervasive across a broad range of personal and social situations.
C.

The enduring pattern leads to clinically significant distress or impairment in social, occupational, or other important
areas of functioning.

D.

The pattern is stable and of long duration, and its onset can be traced back at least to adolescence or early adulthood.

E. The enduring pattern is not better accounted for as a manifestation or consequence of another mental disorder.
F.

The enduring pattern is not due to the direct physiological effects of a substance (e.g., a drug of abuse, a medication) or
a general medical condition (e.g., head trauma).

Source. Reprinted from American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association, 2000. Used with permission. Copyright 2000 American Psychiatric Association.

of inner experience and behavior is manifest by characteristic patterns of 1) cognition (i.e., ways of perceiving
and interpreting self, other people, and events); 2) affectivity (i.e., the range, intensity, lability, and appropriateness of emotional response); 3) interpersonal functioning; and 4) impulse control. Patients with personality
disorders are expected to have manifestations in at least
two of these areas.

Cognitive Features
Personality disorders commonly affect the ways patients think about their relationships with other people
and about themselves. Most of the DSM-IV-TR diagnostic criteria for paranoid personality disorder reflect
a disturbance in cognition, characterized by pervasive
distrust and suspiciousness of others. Patients with
paranoid personality disorder suspect that others are
exploiting, harming, or deceiving them; doubt the loyalty or trustworthiness of others; read hidden, demeaning, or threatening meanings into benign remarks or
events; and perceive attacks on their character or reputation. Among the major symptoms of schizotypal
personality disorder are characteristic cognitive and
perceptual distortions, such as ideas of reference; odd
beliefs and magical thinking (e.g., superstitiousness, belief in clairvoyance or telepathy); bodily illusions; and
suspiciousness and paranoia similar to that observed in
patients with paranoid personality disorder.
Patients with borderline personality disorder (BPD)
may also experience transient paranoid ideation when
under stress, but the characteristic cognitive manifestations of borderline patients are dramatic shifts in their

views toward people with whom they are intensely


emotionally involved. These shifts result in their overidealizing others at one point and then devaluating
them at another point, when they feel disappointed,
neglected, or uncared for. This phenomenon is commonly referred to as splitting. Patients with narcissistic personality disorder exhibit a grandiose sense of
self; have fantasies of unlimited success, power, brilliance, beauty, or ideal love; and believe that they are
special or unique.
Patients with avoidant personality disorder have
excessively negative opinions of themselves, in contrast to patients with narcissistic personality disorder.
They see themselves as inept, unappealing, and inferior, and they constantly perceive that they are being
criticized or rejected. Patients with dependent personality disorder also lack self-confidence and believe
that they are unable to make decisions or to take care
of themselves. Patients with obsessive-compulsive
personality disorder (OCPD) are perfectionistic and
rigid in their thinking and are often preoccupied with
details, rules, lists, and order.

Affective Features
Some patients with personality disorders are emotionally constricted, whereas others are excessively emotional. Among the constricted types are patients with
schizoid personality disorder, who experience little
pleasure in life, appear indifferent to praise or criticism, and are generally emotionally cold, detached,
and unexpressive. Patients with schizotypal personality disorder also often have constricted or inappropri-

Manifestations, Clinical Diagnosis, and Comorbidity

ate affect, although they can exhibit anxiety in relation


to their paranoid fears. Patients with OCPD have considerable difficulty expressing loving feelings toward
others, and when they do express affection, they do so
in a highly controlled or stilted manner.
Among the most emotionally expressive patients
with personality disorders are those with borderline
and histrionic personality disorders. Patients with BPD
are emotionally labile and react very strongly, particularly in interpersonal contexts, with a variety of intensely dysphoric emotions, such as depression, anxiety,
or irritability. They are also prone to inappropriate,
intense outbursts of anger and are often preoccupied
with fears of being abandoned by those they are attached to and reliant upon. Patients with histrionic personality disorder often display rapidly shifting emotions that seem to be dramatic and exaggerated but are
shallow in comparison to the intense emotional expression seen in BPD. Patients with antisocial personality
disorder (ASPD) characteristically have problems with
irritability and aggressive feelings toward others, which
are expressed in the context of threat or intimidation. Patients with narcissistic personality disorder display arrogant, haughty attitudes and have no empathy for other
people. Patients with avoidant personality disorder are
dominated by anxiety in social situations; those with dependent personality disorder are preoccupied by anxiety over the prospects of separation from caregivers and
the need to be independent.

Interpersonal Features
Interpersonal problems are probably the most typical of
personality disorders (Benjamin 1996; Kiesler 1996).
Other mental disorders are characterized by prominent
cognitive or affective features or by problems with impulse control. All personality disorders, however, also
have interpersonal manifestations that can be described
along the two orthogonal poles of the so-called interpersonal circumplex: dominance versus submission
and affiliation versus detachment (Wiggins 2003; see
also Chapter 3, Categorical and Dimensional Models
of Personality Disorders, and Chapter 20, Interpersonal Therapy).
Personality disorders characterized by a need for
or a tendency toward dominance in interpersonal relationships include antisocial, histrionic, narcissistic,
and obsessive-compulsive. Patients with ASPD deceive and intimidate others for personal gain. Patients
with histrionic and narcissistic personality disorders
need to be the center of attention and require excessive
admiration, respectively. Patients with OCPD need to
control others and have them submit to their ways of

59

doing things. On the submissive side are patients with


avoidant and dependent personality disorders. Patients with avoidant personality disorder are inhibited
in interpersonal relationships because they are afraid
of being shamed or ridiculed. Patients with dependent
personality disorder will not disagree with important
others for fear of losing their support or approval and
will actually do things that are unpleasant, demeaning, or self-defeating in order to receive nurturance
from them. Patients with BPD may alternate between
submissiveness and dominance, seeming to become
deeply involved and dependent only to turn manipulative and demanding when their needs are not met.
In the domain of affiliation versus detachment,
patients with histrionic, narcissistic, and dependent
personality disorders have the greatest degrees of
affiliative behavior, whereas patients with paranoid,
schizoid, schizotypal, avoidant, and obsessive-compulsive personality disorders are the most detached. Patients with histrionic, narcissistic, and dependent personality disorders are pro-social because of their needs
for attention, admiration, and support, respectively. Patients with paranoid personality disorder do not trust
others enough to become deeply involved; patients with
schizotypal personality disorder have few friends or
confidants, in part from a lack of trust and in part as a result of poor communication and inadequate relatedness.
Patients with avoidant personality disorder are socially
isolated because of their feelings of inadequacy and their
fears of rejection, whereas those with schizoid personality disorder neither desire nor enjoy relationships. Patients with OCPD opt for work and productivity over
friendships and interpersonal activity because they feel
more in control in the former than the latter. Patients
with BPD again can vacillate between being overly attached and dependent on someone (often one who is not
the best match) and being isolated, distant, and aloof.

Problems With Impulse Control


Problems with impulse control can also be viewed as
extremes on a continuum. Personality disorders characterized by a lack of impulse control include ASPD and
BPD. Disorders involving problems with overcontrol include avoidant, dependent, and obsessive-compulsive
personality disorders. ASPD is a prototype of a personality disorder characterized by impulsivity. Patients
with ASPD break laws, exploit others, fail to plan ahead,
get into fights, ignore commitments and obligations,
and exhibit generally reckless behaviors without regard
to consequences, such as speeding, driving while intoxicated, having impulsive sex, or abusing drugs. Patients
with BPD also show many problems with impulse con-

60

Table 42.
Cluster

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

DSM-IV-TR personality clusters, specific types, and their defining clinical features
Type

Characteristic Features
Odd or eccentric

Paranoid

Pervasive distrust and suspiciousness of others such that their motives


are interpreted as malevolent

Schizoid

Pervasive pattern of detachment from social relationships and


restricted range of expression of emotions in interpersonal settings

Schizotypal

Pervasive pattern of social and interpersonal deficits marked by acute


discomfort with, and reduced capacity for, close relationships as well
as by cognitive or perceptual distortions and eccentricities of
behavior

Dramatic, emotional, or erratic


Antisocial

History of conduct disorder before age 15; pervasive pattern of


disregard for and violation of the rights of others; current age
at least 18

Borderline

Pervasive pattern of instability of interpersonal relationships, selfimage, and affects, and marked impulsivity

Histrionic

Pervasive pattern of excessive emotionality and attention seeking

Narcissistic

Pervasive pattern of grandiosity (in fantasy or behavior), need for


admiration, and lack of empathy

Avoidant

Pervasive pattern of social inhibition, feelings of inadequacy, and


hypersensitivity to negative evaluation

Dependent

Pervasive and excessive need to be taken care of that leads to


submissive and clinging behavior and fears of separation

Obsessive-compulsive

Pervasive pattern of preoccupation with orderliness, perfectionism,


and mental and interpersonal control at the expense of flexibility,
openness, and efficiency

Anxious or fearful

Source. Adapted from American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision.
Washington, DC, American Psychiatric Association, 2000, p. 685. Used with permission. Copyright 2000 American Psychiatric Association.

trol, including impulsive spending, indiscriminate sex,


substance abuse, reckless driving, and binge eating. In
addition, patients with BPD engage in recurrent suicidal
threats, gestures, or attempts and in self-mutilating behavior such as cutting or burning. Finally, patients with
BPD have problems with anger management, have frequent temper outbursts, and at times may even engage
in physical fights.
In contrast, patients with avoidant personality disorder are generally inhibited, especially in relation to
people, and are reluctant to take risks or to undertake
new activities. Patients with dependent personality disorder cannot even make decisions and do not take initiative to start things. Patients with OCPD are overly
conscientious and scrupulous about morality, ethics,
and values; they cannot bring themselves to throw
away even worthless objects and are miserly.
The DSM-IV-TR personality disorder clusters, specific personality disorder types, and their principal defining clinical features are presented in Table 42.

Pervasiveness and Inflexibility


For a personality disorder to be present, the disturbances reviewed earlier have to be manifest frequently
over a wide range of behaviors, feelings, and perceptions and in many different contexts. In DSM-IV-TR,
attempts are made to stress the pervasiveness of the behaviors caused by personality disorders. Added to the
basic definition of each personality disorder, serving as
the stem to which individual features apply, is the
phrase present in a variety of contexts. For example,
the essential features of paranoid personality disorder
in DSM-IV-TR, preceding the specific criteria, begin:
A pervasive distrust and suspiciousness of others
such that their motives are interpreted as malevolent,
beginning by early adulthood and present in a variety
of contexts, as indicated by four (or more) of the following (American Psychiatric Association 2000, p.
694). Similarly, for dependent personality disorder, the
criteria are preceded by the description: A pervasive

Manifestations, Clinical Diagnosis, and Comorbidity

and excessive need to be taken care of that leads to submissive and clinging behavior and fears of separation,
beginning by early adulthood and present in a variety
of contexts, as indicated by five (or more) of the following (American Psychiatric Association 2000, p. 725).
Inflexibility is a feature that helps to distinguish
personality traits or styles and personality disorders.
Inflexibility is indicated by a narrow repertoire of responses that are repeated even when the situation
calls for an alternative behavior or in the face of clear
evidence that a behavior is inappropriate or not working. For example, an obsessive-compulsive person
rigidly adheres to rules and organization even in recreation and loses enjoyment as a consequence. An
avoidant person is so fearful of being scrutinized or
criticized, even in group situations in which he or she
could hardly be the focus of such attention, that life becomes painfully lonely.

Onset and Clinical Course


Personality and personality disorders have traditionally been assumed to reflect stable descriptions of a person, at least after a certain age. Thus, the patterns of inner experience and behaviors described earlier are
called enduring. Personality disorder is also described as of long duration, with an onset that can be
traced back to at least adolescence or early adulthood
(American Psychiatric Association 2000, p. 686). These
concepts persist as integral to the definition of personality disorder despite a large body of empirical evidence
that suggests that personality disorder psychopathology is not as stable as the DSM definition would indicate. Longitudinal studies indicate that personality disorders tend to improve over time, at least from the point
of view of their overt clinical signs and symptoms (Grilo
et al. 2004b; Johnson et al. 2000; Lenzenweger 1999; Shea
et al. 2001). Furthermore, personality disorder criteria
sets consist of combinations of pathological personality
traits and symptomatic behaviors (McGlashan et al. in
press). Some behaviors, such as self-mutilating behavior (BPD), may be evidenced much less frequently than
traits such as views self as socially inept, personally
unappealing or inferior to others (avoidant personality
disorder). How stable individual manifestations of personality disorders actually are and what the stable components of personality disorders are have become areas
of active empirical research. It may be that personality
psychopathology waxes and wanes depending on the
circumstances of a persons life (see Chapter 6, Course
and Outcome of Personality Disorders).

61

Distress or Impairment in Functioning


Another important aspect of personality disorders
that distinguishes them from traits or styles is that personality disorders lead to distress or impairment in
functioning. By their nature, some personality disorders may not be accompanied by obvious subjective
distress on the part of the patient. Examples would include schizoid personality disorder, in which a patient
is ostensibly satisfied with his or her social isolation
and does not seem to need or desire the companionship of others, and ASPD, in which the patient has utter disdain and disregard for social norms and will not
experience distress unless his activities are thwarted.
On the other side of the coin are patients with BPD,
who are likely to experience and express considerable
distress, especially when disappointed in a significant
other, or patients with avoidant personality disorder,
who, in contrast with schizoid patients, are usually
very uncomfortable and unhappy with their lack of
close friends and companions.
All personality disorders are maladaptive, however, and are accompanied by functional problems in
school or at work, in social relationships, or at leisure.
The requirement for impairment in psychosocial functioning is codified in DSM-IV-TR in its criterion C of
the general diagnostic criteria for a personality disorder, which states that the enduring pattern [of inner
experience and behavior, i.e., personality] leads to
clinically significant distress or impairment in social,
occupational, or other important areas of functioning
(American Psychiatric Association 2000, p. 689).
A number of studies have compared patients with
personality disorders to patients with no personality
disorder or with Axis I disorders and have found that
patients with personality disorders were more likely
to be separated, divorced, or never married (Drake
and Vaillant 1985; Pfohl et al. 1984; Zimmerman and
Coryell 1989) and to have had more unemployment,
frequent job changes, or periods of disability (McGlashan 1986; Modestin and Villiger 1989; Paris et al.
1987; Swartz et al. 1990). It is interesting that only
rarely have patients with personality disorders been
found to be less well educated (Reich et al. 1989; Soloff
and Ulrich 1981). Fewer studies have examined quality of functioning, but in those that have, poorer social
functioning or interpersonal relationships (Noyes et
al. 1990; Torgersen 1984; Turner et al. 1991) and poorer
work functioning or occupational achievement and
satisfaction have been found among patients with personality disorders than with others (Andreoli et al.
1989; Casey and Tyrer 1990; Pope et al. 1983; Shea et al.

62

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

1990). When patients with different personality disorders have been compared with each other on levels of
functional impairment, those with severe personality
disorders such as schizotypal and borderline have
been found to have significantly more impairment at
work, in social relationships, and at leisure than patients with less severe personality disorders, such as
OCPD, or with an impairing Axis I disorder, such as
major depressive disorder (MDD) without personality
disorder. Patients with avoidant personality disorder
had intermediate levels of impairment. Even the less
impaired patients with personality disorders (i.e.,
OCPD), however, had moderate to severe impairment
in at least one area of functioning (or a Global Assessment of Functioning rating of 60 or less) (Skodol et al.
2002). The finding that significant impairment may be
in only one area suggests that patients with personality disorders differ not only in the degree of associated
functional impairment but also in the breadth of impairment across functional domains.
Another important aspect of the impairment in
functioning in patients with personality disorders is
that it tends to be persistent even beyond apparent improvement in personality disorder psychopathology
itself (Seivewright et al. 2004; Skodol et al. in press).
The persistence of impairment is understandable if
one considers that personality disorder psychopathology has usually been long-standing and, therefore, has
disrupted a persons work and social development
over a period of time (Roberts et al. 2003). The scars
or residua of personality disorder pathology take time
to heal or be overcome. With time (and treatment),
however, improvements in functioning can occur.

APPROACHES TO CLINICAL INTERVIEWING


Interviewing a patient to assess for a possible personality disorder presents certain challenges that are
somewhat unique. Thus, the interviewer is likely to
need to rely on a variety of techniques for gathering
information to arrive at a clinical diagnosis, including
observation and interaction with the patient, direct
questioning, and interviewing informants.

Observation and Interaction


One problem in evaluating a patient for a personality
disorder arises from the fact that most people are not
able to view their own personality objectively (Zimmerman 1994). Because personality is, by definition,
the way a person sees, relates to, and thinks about

himself or herself and the environment, a persons assessment of his or her own personality must be colored by it. The expression of Axis I psychopathology
may also be colored by Axis II personality stylefor
example, symptoms exaggerated by the histrionic or
minimized by the compulsive personalitybut the
symptoms of Axis I disorders are usually more clearly
alien to the patient and more easily identified as problematic. People usually learn about their own problem
behavior and their patterns of interaction with others
through the reactions or observations of other people
in their environments.
Traditionally, clinicians have not conducted the
same kind of interview in assessing patients suspected
of having a personality disturbance as they do with
persons suspected of having, for example, a mood or
an anxiety disorder. Rather than directly questioning
the patient about characteristics of his or her personality, the clinician, assuming that the patient cannot accurately describe these traits, looks for patterns in the
way the patient describes social relations and work
functioning. These two areas usually give the clearest
picture of personality style in general and personality
problems specifically. Clinicians have also relied
heavily on their observations of how patients interact
with them during an evaluation interview or in treatment as manifestations of their patients personalities
(Westen 1997).
These approaches have the advantage of circumventing the lack of objectivity patients might have
about their personalities, but they also create problems.
The clinician usually comes away with a global impression of the patients personality but frequently is not
aware of many of that patients specific personality
characteristics because he or she has not made a systematic assessment of the signs and symptoms of the
wide range of personality disorders (Blashfield and
Herkov 1996; Morey and Ochoa 1989; Zimmerman and
Mattia 1999b). In routine clinical practice, clinicians
tend to use the nonspecific DSM-IV-TR diagnosis of
PDNOS when they believe that a patient meets the general criteria for a personality disorder, because they often do not have enough information to make a specific
diagnosis (Widiger and Saylor 1998). Alternatively, clinicians will diagnose personality disorders hierarchically: once a patient is seen as having one (usually
severe) personality disorder, the clinician will not assess whether traits of other personality disorders are
present (Adler et al. 1990; Herkov and Blashfield 1995).
Reliance on interaction with the clinician for personality diagnosis runs the risk of generalizing a mode
of interpersonal relating that may be limited to a par-

Manifestations, Clinical Diagnosis, and Comorbidity

ticular situation or contextthat is, the evaluation itself. Although the interaction of patient and clinician
can be a useful and objective observation, caution
should be used in interpreting its significance, and attempts must be made to integrate this information
into a broader overall picture of patient functioning.

63

interview is useful clinically when the results of an assessment might be subject to close scrutiny, such as in
child custody, disability, or forensic evaluations (Widiger and Coker 2002). Instruments to assist the clinician in the assessment of personality psychopathology are presented in detail in Chapter 5, Assessment
Instruments and Standardized Evaluation.

Direct Questioning
Interviewing Informants
In psychiatric research, a portion of the poor reliability of personality disorder diagnosis has been assumed to be due to the variance in information resulting from unsystematic assessment of personality
traits. Therefore, efforts have been made to develop
various structured methods for assessing personality
disorders (Kaye and Shea 2000) comparable with
those that have been successful in reducing information variance in assessing Axis I disorders (Skodol
and Bender 2000). These methods include both 1)
self-report measures such as the Personality Disorders Questionnaire4 (Hyler 1994), the Millon Clinical Multiaxial InventoryIII (Millon et al. 1997), and
the Minnesota Multiphasic Personality Inventory2
(Somwaru and Ben-Porath 1995); and 2) clinical interviews such as the Structured Interview for DSM-IV
Personality Disorders (Pfohl et al. 1997), the International Personality Disorder Examination (Loranger
1999), the Structured Clinical Interview for DSM-IV,
Axis II (First et al. 1997), the Diagnostic Interview for
DSM-IV Personality Disorders (Zanarini et al. 1996),
and the Personality Disorder InterviewIV (Widiger
et al. 1995).
The interviews are based on the general premise
that the patient can be asked specific questions that
will indicate the presence or absence of each of the criteria of each of the 10 DSM-IV-TR personality disorder types. The self-report instruments are generally
considered to require a follow-up interview because
of a very high rate of apparently false-positive responses, but data from studies comparing self-report
measures with clinical interviews suggest that the
former aid in identification of personality disturbances (Hyler et al. 1990, 1992). Thus, the clinician can
keep in mind that patients do not necessarily deny
negative personality attributes: in fact, the evidence
suggests that they may even overreport traits that clinicians might not think are very important, and that
patients can, if asked, consistently describe a wide
range of personality traits to multiple interviewers. A
self-report inventory might be an efficient way to help
focus a clinical interview on a narrower range of personality disorder psychopathology. A semistructured

Frequently, a patient with a personality disorder consults a mental health professional for evaluation or
treatment because another person has found his or her
behavior problematic. This person may be a boss,
spouse, boyfriend or girlfriend, teacher, parent, or
representative of a social agency. Indeed, some people
with personality disorders do not even recognize the
problematic aspects of their manner of relating or perceiving except as it has a negative effect on someone
with whom they interact.
Because of these blind spots that people with personality disorders may have, the use of a third-party informant in the evaluation can be useful (Zimmerman et
al. 1986). In some treatment settings, such as a private
individual psychotherapy practice, it may be considered counterproductive or contraindicated to include a
third party, but in many inpatient and outpatient settings, certainly during the evaluation process, it may be
appropriate and desirable to see some person close to
the patient to corroborate both the patients report and
ones own clinical impressions.
Of course, there is no reason to assume that the
informant is bias-free or not coloring a report about
the patient with his or her own personality style. In
fact, the correspondence between patient self-assessments of personality disorder psychopathology and
informant assessments has been generally found to
be modest at best (Klonsky et al. 2002). Agreement
on pathological personality traits, temperament, and
interpersonal problems appears to be somewhat better than on DSM personality disorders. Informants
usually report more personality psychopathology
than patients. Self/informant agreement on personality disorders is highest for Cluster B disorders (excluding narcissistic personality disorder), lower for
Clusters A and C, and lowest for traits related to narcissism and entitlement, as might be expected. So the
clinician must make a judgment about the objectivity of the informant and use this as a part, but not a
sufficient part, of the overall data on which to base a
personality disorder diagnosis (Zimmerman et al.
1988). Which source, the patient or the informant,

64

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

provides information that is more useful for clinical


purposes, such as choosing a treatment or predicting
outcome (e.g., Klein 2003), is yet to be definitively
determined.

PROBLEMS IN CLINICAL ASSESSMENT


Assessing Pervasiveness
The pervasiveness of personality disturbance can be
difficult to determine. When a clinician inquires if a
person often has a particular experience, a patient
will frequently reply sometimes, which then has to
be judged for clinical significance. What constitutes a
necessary frequency for a particular trait or behavior
(Widiger 2002) and in how many different contexts or
with how many different people the trait or behavior
needs to be expressed has not been well worked out.
Clinicians are forced to rely on their own judgment,
keeping in mind also that maladaptivity and inflexibility are hallmarks of pathological traits.
For the clinician interviewing a patient with a possible personality disorder, data about the many areas
of functioning, the interpersonal relationships with
people interacting in different social roles with the patient, and the nature of the patientclinician relationship should be integrated into a comprehensive assessment of pervasiveness. Too often, clinicians place
disproportionate importance on a patients functioning at a particular job or with a particular boss or significant other person.

State Versus Trait


An issue that cuts across all personality disorder diagnoses and presents practical problems in differential
diagnosis is the distinction between clinical state and
personality trait. Personality is presumed to be an enduring aspect of a person, yet assessment of personality ordinarily takes place cross-sectionallythat is,
over a brief interval in time. Thus, the clinician is challenged to separate out long-term dispositions of the
patient from other more immediate or situationally
determined characteristics. This task is more complicated by the fact that the patient often comes for evaluation when there is some particularly acute problem,
which may be a social or job-related crisis or the onset
of an Axis I disorder (Shea 1997). In either case, the situation in which the patient is being evaluated is frequently a state that is not completely characteristic of
the patients life over the longer run.

Assessing an Enduring Pattern


DSM-IV-TR indicates that personality disorders are of
long duration and are not better accounted for as a
manifestation or consequence of another mental disorder (American Psychiatric Association 2000, p. 689).
Making these determinations in practice is not easy.
First of all, an accurate assessment requires recognition of current state. An assessment of current state, in
turn, includes knowledge of the circumstances that
have prompted the person to seek treatment, the consequences in terms of the decision to seek treatment,
the current level of stress, and any actual Axis I psychopathology, if present.
The DSM-IV-TR multiaxial system is of considerable aid in the assessment of these problems because
of its separation of Axis I disorders from Axis II disorders and its individual axes for physical disorders and
psychosocial stressors. A multiaxial system forces clinicians to think about the effects of aspects of patients
current state on long-term patterns of behavior, but it
does not make the distinctions for them.
It is not clear from the diagnostic criteria of DSMIV-TR how long a pattern of personality disturbance
needs to be present, or when it should become evident, for a personality disorder to be diagnosed. Earlier iterations of the DSM stated that patients were
usually 18 years of age or older because it can be argued that, up to that age, a personality pattern could
neither have been manifest long enough nor have become significantly entrenched to be considered a stable constellation of behavior. DSM-IV-TR states, however, that some manifestations of personality disorder
are usually recognizable by adolescence or earlier and
that personality disorders can be diagnosed in individuals younger than 18 years if manifestations are
present for at least 1 year. Longitudinal research has
shown that personality disorder symptoms evident in
childhood or early adolescence may not persist into
adult life (Johnson et al. 2000). Longitudinal research
has also shown that there is continuity between certain disorders of childhood and adolescence and personality disorders in early adulthood (Kasen et al.
1999, 2001). Thus, a young boy with oppositional
defiant or attention-deficit/hyperactivity disorder in
childhood may go on to develop conduct disorder as
an adolescent, which can progress to full-blown ASPD
in adulthood (Bernstein et al. 1996; Lewinsohn et al.
1997; Rey et al. 1995; Zoccolillo et al. 1992). ASPD is the
only diagnosis not given before age 18; an adolescent
exhibiting significant antisocial behavior before age 18
is diagnosed with conduct disorder.

Manifestations, Clinical Diagnosis, and Comorbidity

Regarding the course of a personality disorder,


DSM-IV-TR states that personality disorders are relatively stable over time, although certain of them (e.g.,
ASPD and BPD) may become somewhat attenuated
with age, whereas others may not or may, in fact, become more pronounced (e.g., obsessive-compulsive
and schizotypal personality disorders). As mentioned
earlier and discussed in greater detail in Chapter 6,
Course and Outcome of Personality Disorders, this
degree of stability may not necessarily pertain to all of
the features of all DSM-IV-TR personality disorders
equally.
To assess stability retrospectively, the clinician must
ask questions about periods of a persons life that are of
various degrees of remoteness from the current situation. Retrospective reporting is subject to distortion,
however, and the only sure way of demonstrating stability over time is, therefore, to do prospective follow-up evaluations. Thus from a practical, clinical point
of view, personality disorder diagnoses made crosssectionally and on the basis of retrospectively collected
data would be tentative or provisional pending confirmation by longitudinal evaluation. On an inpatient
service, a period of intense observation by many professionals from diverse perspectives may suffice to establish a pattern over time (Skodol et al. 1988, 1991). In
a typical outpatient setting in which there are much less
frequent encounters with a patient, more time may be
required. Ideally, features of a personality disorder
should be evident over years, but it is not practical to
delay inordinate amounts of time before coming to a
diagnostic conclusion. A good retrospective history
confirmed by a period of prospective evaluation should
make the personality pattern evident.

Assessing the Effect of Axis I Disorder


An Axis I disorder can complicate the diagnosis of a
personality disorder in several ways (Widiger and
Sanderson 1995; Zimmerman 1994). An Axis I disorder may cause changes in a persons behavior or attitudes that can appear to be signs of a personality disorder. Depression, for example, may cause a person to
seem excessively dependent, avoidant, or self-defeating. Cyclothymia or bipolar disorder (not otherwise
specified; bipolar II) may lead to periods of grandiosity, impulsivity, poor judgment, and depression that
might be confused with manifestations of narcissistic
or borderline personality disorders.
The clinician must be aware of the Axis I psychopathology and attempt to assess Axis II independently. This assessment can be attempted in one of two
ways. First, the clinician can ask about aspects of per-

65

sonality functioning at times when the patient is not


experiencing Axis I symptoms. This approach is feasible when the Axis I disorder is of recent onset and
short duration or, if more chronic, if the course of the
disorder has been characterized by relatively clear-cut
episodes with complete remission and symptom-free
periods of long duration. When the Axis I disorder is
chronic and unremitting, then the Axis I psychopathology and personality functioning blend together to
an extent that makes differentiating between them
clearly artificial.
A second approach to distinguishing signs of
Axis I pathology from signs of Axis II personality is
longitudinal and would defer an Axis II diagnosis
pending the outcome of a trial of treatment for the
Axis I disorder. This strategy may be the preferred approach in the case of a long-standing and chronic Axis
I disorder, like cyclothymia, that has never been previously recognized or treated. Although one always
runs the risk of a partial response to treatment and
some residual symptoms, this tactic may bring the clinician as close, practically speaking, as he or she will
get to observing the patients baseline functioning.

Case Example
The following case is adapted from Skodol (1989).
A 24-year-old, unemployed man sought psychiatric
hospitalization because of a serious problem with
depression. The man reported that he had felt mildly,
but continuously, depressed since the age of 16.
When he reached his twenties, he had begun to have
more severe bouts that made him suicidal and unable to function.
During the most recent episode, beginning about
6 months previously, he had quit his job as a taxi
driver and isolated himself from his friends. He
spent his time lying around and eating a lot and, in
fact, had gained 60 pounds. He had difficulty falling
asleep, felt fatigued all day long, could not concentrate, felt worthless (Theres no purpose to my life)
and guilty (I missed my chances; Ive put my family
through hell), and had taken an overdose of sleeping pills.
The man received a semistructured interview assessment of Axis II psychopathology. In describing
his personality, he said that he once thought of himself as lively and good-natured, but that over the
past 4 or 5 years, he felt he had changed. He said that
he was very sensitive to criticism, afraid to get involved with people, fearful of new places and experiences, convinced he was making a fool of himself,
and afraid of losing control. He felt very dependent
on others for decision making and for initiative. He
said that he was so needy of others that they
could do anything to him and he would take it.

66

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

He felt helpless when alone, was sure he would end


up alone and in the streets, and was constantly
looking to others, especially family members, for
comfort and reassurance.
The man also thought that people took advantage of him now and that he let them because he
never stood up for his own self-interest. He felt like
a total failure with no redeeming virtues. He said he
either deliberately passed up opportunities to improve his situation because he felt I dont deserve
any better or else undermined himself without
thinking by failing to follow through, for example,
on a job interview. He believed that no one could really be trusted, that old friends probably talked
about him behind his back (They think Im a slob),
that he could not open up with new people because
they too would eventually turn on him and reject
him, and that he now carried a chip on his shoulder
because he had been burned by others so often. He
admitted that he was not blame-free in relationships
because he had also used people, especially members of his family.
The patient felt that he was not improving in his
outpatient treatment of the last 3 years. His reason
for seeking hospitalization, in addition to the fact
that he continually thought of suicide and was
frightened he might actually succeed in killing himself, was that he felt totally lost in his life, without
direction, goals, or knowing what mattered to him.
He said he felt hollow. If they cut me open after
I was dead, he said, theyd probably find out I was
all shriveled up inside.
This mans description of his personality, the
ways in which he characteristically thought about
himself, saw others and his relationships to them, and
behaved, actually met DSM-IV-TR criteria for avoidant, dependent, paranoid, and borderline personality
disorders. He was hospitalized for long-term treatment, which was available at the time. In addition to
receiving individual, psychoanalytically oriented psychotherapy sessions and participating in a variety of
therapeutic groups, he was given fluoxetine, up to 80
mg/day, for treatment of Axis I MDD and dysthymia.
Six months after admission, the patient reported
that he felt significantly less depressed. Measured in
terms of the Hamilton Rating Scale for Depression,
the initial severity of his depression was 30, and his
posttreatment score was 10. A repeat semistructured
assessment of his personality functioning revealed
that he no longer met DSM-IV-TR criteria for any
personality disorder, although he continued to exhibit some dependent traits.

Another example of the way in which Axis I and II


disorders interact to obscure differential diagnosis is
the case of apparent Axis II psychopathology that, in
fact, is the prodrome of an Axis I disorder. Distinguishing Cluster A personality disorders, such as
paranoid, schizoid, and schizotypal, from the early
signs of Axis I disorders in the schizophrenia and

other psychotic disorders class can be particularly difficult. If a clinician is evaluating a patient early in the
course of the initial onset of a psychotic disorder, he or
she may be confronted with changes in the person toward increasing suspiciousness, social withdrawal,
eccentricity, or reduced functioning. Because the diagnosis of psychotic disorders, including schizophrenia,
requires that the patient have an episode of active psychosis with delusions and hallucinations, it is not possible to diagnose this prodrome as a psychotic disorder. In fact, until the full-blown disorder is present, the
clinician cannot be certain if it is, indeed, a prodrome.
If a change in behavior is of recent onset, then it
does not meet the stability criteria for a personality disorder. In such cases, the clinician is forced to diagnose
an unspecified mental disorder (nonpsychotic; DSMIV-TR code 300.9). If, however, the pattern of suspiciousness or social withdrawal with or without eccentricities has been well established, it may legitimately
be a personality disorder and be diagnosed as such.
If the clinician follows such a patient over time and
the patient develops a full-fledged psychotic disorder,
the personality disturbance is no longer adequate for a
complete diagnosis because none of the Axis II disorders includes frankly psychotic symptoms. This fairly
obvious point is frequently overlooked in practice. All
of the personality disorders that have counterpart psychotic disorders on Axis I have milder symptoms in
which reality testing is, at least in part, intact. For instance, a patient with paranoid personality disorder
may have referential ideas but not frank delusions of
reference, and a patient with schizotypal personality
disorder may have illusions but not hallucinations. A
possible exception is BPD, in which brief psychotic experiences (lasting minutes to an hour or two at most)
are included in the diagnostic criteria. In all cases, however, when the patient becomes psychotic for even a
day or two, an additional Axis I diagnosis is necessary.
For the patient with a diagnosis of schizotypal personality disorder, the occurrence of a psychotic episode
of 1 months duration almost certainly means the disturbance will meet the criteria for schizophrenia, the
symptoms of schizotypal personality disorder counting as prodromal symptoms toward the 6-month duration requirement. Under these circumstances, the diagnosis of schizophrenia, with its pervasive effects on
cognition, perception, functional ability, and so on, is
sufficient, and a diagnosis of schizotypal personality
disorder is redundant. When the patient becomes nonpsychotic again, he or she would be considered to have
residual schizophrenia instead of schizotypal personality disorder.

Manifestations, Clinical Diagnosis, and Comorbidity

Personality Traits Versus Personality


Disorders
Another difficult distinction is between personality traits
or styles and personality disorders. All patientsall people for that mattercan be described in terms of distinctive patterns of personality, but all do not necessarily
warrant a diagnosis of personality disorder. This error is
particularly common among inexperienced evaluators.
The important features that distinguish pathological personality traits from normal traits are their inflexibility
and maladaptiveness, as discussed earlier.
DSM-IV-TR recognizes that it is important to describe personality style as well as to diagnose personality disorder on Axis II. Therefore, instructions are included to list personality features on Axis II even
when a personality disorder is absent, or to include
them as modifiers of one or more diagnosed personality disorders (e.g., BPD with histrionic features). In
practice, however, this option has been seldom utilized (Skodol et al. 1984), even though research has
shown that, in addition to the approximately 50% of
clinic patients who meet criteria for a personality disorder, another 35% warrant information descriptive of
their personality styles on Axis II (Kass et al. 1985).
The overlap among the features of personality disorders also becomes very evident when emphasis is
placed on the assessment of traits of all personality
disorders, even when one is predominant.
The following case example describes a patient
with an Axis I disorder whose ongoing treatment was
very much affected by Axis II personality traits, none
of which met criteria for a personality disorder.

Case Example
The following case is adapted from Skodol (1989).
A 25-year-old, single female receptionist was referred for outpatient therapy following hospitalization for her first manic episode. The patient had attended college for 1 year but dropped out in order to
go into advertising. Over the next 5 years, she had
held a series of receptionist, secretarial, and sales
jobs, each of which she quit because she wasnt getting ahead in the world. She lived in an apartment
on the north side of Chicago, by herself, that her parents had furnished for her. She ate all of her meals,
however, at her mothers house and claimed not
even to have a box of crackers in her cupboard. Between her jobs, her parents paid her rent.
Her career problems stemmed from the fact
that, although she felt quite ordinary and without talent for the most part, she had fantasies of a career as
a movie star or high fashion model. She took acting

classes and singing lessons but had never had even a


small role in a play or show. What she desired was
not so much the careers themselves but the glamour
associated with them. Although she wanted to move
in the circles of the beautiful people, she was certain that she had nothing to offer them. She sometimes referred to herself as nothing but a shell and
scorned herself because of it. She was unable to picture herself working her way up along any realistic
career line, feeling both that it would take too long
and that she would probably fail.
She had had three close relationships with men
that were characterized by an intense interdependency that initially was agreeable to both parties. She
craved affection and attention and fell deeply in love
with these men. Eventually, however, she became
overtly self-centered, demanding, and manipulative,
and the man would break off the relationship. After
breaking up, she would almost immediately start
claiming that the particular man was going nowhere, was not for her, and would not be missed. In
between these relationships, she often had periods in
which she engaged in a succession of one-night
stands, having sex with a half-dozen partners in a
month. Alternatively, she would frequent rock clubs
and bars, in-spots, as she called them, merely on
the chance of meeting someone who would introduce her to the glamorous world she dreamed of.
The patient had no female friends other than her
sister. She could see little use for such friendships.
She preferred spending her time shopping for stylish
clothes or watching television alone at home. She
liked to dress fashionably and seductively but often
felt that she was too fat or that her hair was the
wrong color. She had trouble controlling her weight
and would periodically go on eating binges for a few
days that might result in a 10-pound weight gain.
She read popular novels but had very few other interests. She admitted she was bored much of the time
but would not admit that cultural or athletic pursuits
were other than a waste of time.
This patient was referred for outpatient follow-up without an Axis II personality disorder diagnosis. In fact, her long-term functioning failed to
meet DSM-IV-TR criteria for any specific type of personality disorder. On the other hand, she almost met
the criteria for several, especially BPD: the patient
showed signs of impulsivity (overeating, sexual promiscuity), intense interpersonal relationships (manipulative, overidealization/devaluation), identity
disturbance, and chronic feelings of emptiness. She
did not, however, display intense anger, intolerance
of being alone, physically self-damaging behavior,
stress-related paranoia or dissociation, or affective
instability independent of her mood disorder. Similarly, she had symptoms of histrionic personality disorder: she was inappropriately sexually seductive
and used her physical appearance to draw attention
to herself, but she was not emotionally overdramatic. She had shallow expression of emotions and
was uncomfortable when she was not the center of
attention, but was not overly suggestible. She also

67

68

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

had some features of narcissistic, avoidant, and dependent personality disorders. The attention paid to
personality traits in her outpatient clinic evaluation
conveyed a vivid picture of the patients complicated
personality pathology, which became the focus of
her subsequent therapy.

Effects of Gender, Culture, and Age


Gender
Although definitive estimates about the sex ratio of
personality disorders cannot be made because ideal
epidemiological studies do not exist, some personality
disorders are believed to be more common in clinical
settings among men and others among women. Those
listed in DSM-IV-TR as occurring more often among
men are paranoid, schizoid, schizotypal, antisocial,
narcissistic, and obsessive-compulsive personality disorders. Those occurring more often in women are borderline, histrionic, and dependent personality disorders. Avoidant is said to be equally common in men
and women. Apparently elevated sex ratios that do not
reflect true prevalence rates can be the result of sampling or diagnostic biases in clinical settings (Widiger
1998). Factors affecting the sex ratios of personality disorders are addressed in detail in Chapter 34, Gender.

Culture
Apparent manifestations of personality disorders must
be considered in the context of a patients cultural reference group and the degree to which behaviors such
as diffidence, passivity, emotionality, emphasis on
work and productivity, and unusual beliefs and rituals
are culturally sanctioned. Only when such behaviors
are clearly in excess or discordant with the standards
of a persons cultural milieu would the diagnosis of a
personality disorder be considered. Certain sociocultural contexts may lend themselves to eliciting and
reinforcing behaviors that might be mistaken for personality disorder psychopathology. Members of minority groups, immigrants, or refugees, for example,
might appear overly guarded or mistrustful, avoidant,
or hostile in response to experiences of discrimination,
language barriers, or problems in acculturation. Cultural issues relevant to the diagnosis and treatment of
personality disorders are the subjects of Chapter 35,
Cross-Cultural Issues.

Age
As mentioned earlier, although personality disorders
are usually not diagnosed prior to the age of 18 years,
certain thoughts, feelings, and behaviors suggestive of
personality psychopathology may be apparent during

childhood. Dependency, social anxiety and hypersensitivity, disruptive behavior, or identity problems, for
example, may be developmentally expected. Followup studies of children have shown decreases in such
behaviors over time (Johnson et al. 2000), although
children with elevated rates of personality disorder
type signs and symptoms do appear to be at higher
risk for both Axis I and Axis II disorders in young
adulthood (Johnson et al. 1999; Kasen et al. 1999).
Thus, some childhood problems may not turn out to
be transitory, and personality disorder may be viewed
developmentally as a failure to mature out of certain
age-appropriate or phase-specific feelings or behaviors. A developmental perspective on personality disorders is presented more fully in Chapter 11, Developmental Issues.

Other Aspects of Personality Functioning


A problem with the DSM conceptualizations of personality disorders is that the individual categories do not
correspond well with existing treatment approaches.
Thus, whether a clinician is a psychodynamically oriented therapist, a cognitive-behavioral therapist, or a
psychopharmacologist, information in addition to that
necessary for a DSM personality disorder diagnosis is
needed to formulate a treatment plan. Usually, this additional information is based on the theory of why a
patient has a personality disorder and/or the mechanisms responsible for perpetuating the dysfunctional
patterns.

Conflicts, Ego Functions, Object Relations, and


Defense Mechanisms
Psychodynamically oriented clinicians have expressed
dissatisfaction with the DSM system of axes, including
Axis II, since its inception. The DSM multiaxial system
fails, in their opinion, to discriminate between patients
according to clinical variables important for planning
treatment with psychodynamic psychotherapy (Karasu
and Skodol 1980). Thus, they may be more interested in
exploring conflicts between wishes, fears, and moral
standards; ego functions such as impulse control or affect regulation; or self and other (object) representations
based on early attachment experiences than on the
signs and symptoms of personality disorders. Elaborations of psychodynamic theories of personality disorders can be found in Chapter 2, Theories of Personality
and Personality Disorders; Chapter 16, Psychoanalysis; and Chapter 17, Psychodynamic Psychotherapies; along with discussions of relevant clinical variables.

Manifestations, Clinical Diagnosis, and Comorbidity

Several groups of researchers (Bond and Vaillant


1986; Perry and Cooper 1989; Vaillant et al. 1986) have
been able to document empirically the clinical utility
of categorizing a patients defensive functioning. Defense mechanisms are automatic psychological processes that protect people against anxiety and against
awareness of internal or external stressors or dangers.
Although this work was considered too early in its development to justify including a separate official axis
based on it, Appendix B in DSM-IV (Criteria Sets and
Axes Provided for Further Study) includes a Defensive Functioning Scale and a Glossary of Specific Defense Mechanisms and Coping Styles. The 27 defense
mechanisms defined in this glossary are acting-out,
affiliation, altruism, anticipation, autistic fantasy, denial, devaluation, displacement, dissociation, helprejecting complaining, humor, idealization, intellectualization, isolation of affect, omnipotence, passive aggression, projection, projective identification, rationalization, reaction formation, repression, self-assertion,
self-observation, splitting, sublimation, suppression,
and undoing. Some defense mechanisms, such as projection, splitting, or acting-out, are always maladaptive, whereas others, such as sublimation or humor,
are adaptive. Patients with personality disorders have
characteristic predominant defensive patterns. Thus
patients with paranoid personality disorder use denial
and projection, those with BPD typically rely on acting-out and splitting (among others), and those with
OCPD use isolation of affect and undoing. Clinicians
may note current defenses or coping styles as well as
a patients predominant current defense level using
the Defensive Functioning Scale. Defensive functioning in patients with personality disorders is the topic
of Chapter 33, Defensive Functioning.

Coping Styles
Although defense mechanisms in DSM-IV-TR are said
to include coping styles, the literature on coping discusses styles not included in the DSM list. Coping refers
to specific thoughts and behaviors that a person uses to
manage the internal and external demands of situations
appraised as stressful (Folkman and Moskowitz 2004;
Lazarus and Folkman 1984; Pearlin and Schooler 1978).
Coping involves cognitive, behavioral, and emotional
responses and may or may not be consistent across
stressful situations or functional roles. Two major broad
styles of coping are problem-focused coping and emotion-focused coping. Problem-focused coping refers to efforts to resolve a threatening problem or diminish its
impact by taking direct action. Emotion-focused coping

69

refers to efforts to reduce the negative emotions aroused


in response to a threat by changing the way the threat is
attended to or interpreted. Meaning-focused and social
coping are other observed coping strategies. Coping has
traditionally been assessed by retrospective self-report
measures (e.g., the Coping Responses Inventory [Moos
1993], the Ways of Coping Questionnaire [Folkman and
Lazarus 1988], and the COPE Inventory [Carver et al.
1989])and more recently by ecological momentary assessment (real-time) techniques (Stone et al. 1998); but
the major types of coping, such as problem solving,
seeking support, distancing and distracting, accepting
responsibility, positive reappraisal, or self-blame, can
also be assessed by clinical interview.

Cognitive Schemas
Cognitive therapists want to characterize patients with
personality disorders according to patients dysfunctional cognitive schemas (core beliefs by which they
process information) or their automatic thoughts,
interpersonal strategies, and cognitive distortions.
Again, particular personality disorders tend to have
particular core beliefs. For example, patients with BPD
frequently have beliefs such as I am needy and weak
or I am helpless if left on my own, whereas patients
with OCPD believe It is important to do a perfect job
on everything or People should do things my way
(Beck et al. 2004). In contrast to beliefs, which map onto
personality disorders specifically, schemas are broader
themes regarding the self and relationships with others
and can cut across personality disorder categories. For
example, a schema of impaired limits can encompass the entitlement of narcissistic personality disorder
as well as the lack of self-control of ASPD or BPD. A
system for assessing and characterizing cognitive schemas and dysfunctional beliefs is included in Chapter
18, Schema Therapy.

Objective Behaviors Versus Inferential Traits


Another difficulty in diagnosing personality disorders
stems from the degree of inference and judgment necessary to make many of the diagnoses. Numerous critics
have noted that it is easy to disagree about symptoms
such as affective instability, self-dramatization, shallow
emotional expression, exaggerated fears, or feelings of
inadequacyall symptoms of DSM-IV-TR personality
disorders. Only the antisocial criteria, among the personality disorders, have historically yielded acceptable levels of reliability, and those criteria have emphasized overtly criminal and delinquent acts.

70

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

These observations led several investigators to attempt to determine sets of behaviors that might serve
to identify types of personality disorder. Although
any one behavior might not be sufficient to indicate a
particular personality trait, multiple behavioral indicators considered together would increase confidence
in recognizing the trait.
Behaviors that typify a particular personality style
have been referred to as prototypical. Livesley (1986) developed a set of prototypical behaviors for the DSM-III
(American Psychiatric Association 1980) personality
disorders and compared them with prototypical traits.
He found that highly prototypical behaviors could be
derived from corresponding traits. For example, with
regard to the concepts of social awkwardness and
withdrawal of the schizoid personality disorder, Livesley found that behaviors such as does not speak unless spoken to, does not initiate social contacts, and
rarely reveals self to others were uniformly rated as
highly prototypic. Corresponding to the overly dramatic and emotional traits of the histrionic personality
disorder were behaviors such as expressed feelings in
an exaggerated way, considered a minor problem
catastrophic, and flirted with several members of the
opposite sex. Behaviors such as has routine schedules and is upset by deviations, overreacted to criticism, and spent considerable time on the minutest
details corresponded to the controlled, perfectionist
traits of OCPD.
DSM-IV-TR makes strides in translating the characteristic traits of the personality disorders into explicit behaviors. The criteria for each personality disorder begin with the definition of the overall style or
set of traits, followed by a listing of ways this might be
expressed. In some instances, for example, for dependent personality disorder, the criteria are quite behavioral. For dependent personality disorder, a pervasive
and excessive need to be taken care of that leads to
submissive and clinging behavior and fears of separation is indicated by such items as has difficulty making everyday decisions without an excessive amount
of advice and reassurance from others and needs
others to assume responsibility for most major areas of
his or her life (American Psychiatric Association
2000, p. 725). For other disorders, such as OCPD, an
example of the behavior is given along with the trait.
For OCPD, perfectionism is indicated by the following
criterion: Shows perfectionism that interferes with
task completion, e.g., is unable to complete a project
because his or her own overly strict standards are not
met (American Psychiatric Association 2000, p. 729).
Not all of the DSM-IV-TR personality disorders are
equally well defined or illustrated by prototypical be-

haviors. Yet because it seems likely that such behaviors are much more reliably recognized than more abstract and inferential traits, the clinician should make
special efforts to elicit examples of behaviors, from patients or other informants, that would constitute objective evidence of the presence of particular personality
traits. Such an approach to assessment is likely to result in more accurate diagnosis.

COMORBIDITY
Since the introduction of a multiaxial system for recording diagnoses in DSM-III, which provided for the
diagnosis of personality disorders on an axis (II) separate from the majority of other mental disorders, it has
become apparent that most patients with personality
disorders also meet criteria for other disorders. Rates
have ranged from about two-thirds to almost 100%
(Dolan-Sewell et al. 2001). The co-occurrence of Axis I
and Axis II disorders has often been referred to as comorbidity, although our current understanding of the
fundamental nature of most mental disorders is insufficient to justify the use of the term according to its formal definition, which requires that a comorbid disorder be distinct from the index disease or condition
(Feinstein 1970). The DSM system, with its tendency to
split as opposed to lump psychopathology via its
many and expanding lists of disorders, encourages the
diagnosis of multiple putative disorders to describe a
patients psychopathology and virtually ensures that
patients will receive more than one diagnosis. In addition to the co-occurrence of personality disorders with
Axis I disorders, it is also common for patients to receive more than one personality disorder diagnosis to
fully describe their personality problems (Lilienfeld et
al. 1994; Oldham et al. 1992). In the sections that follow, major patterns of personality disorder comorbidity will be described.

Co-Occurrence of Personality Disorders


and Axis I Disorders
There are a number of explanations for the high rates
of co-occurrence of personality disorders and Axis I
disorders (Lyons et al. 1997). Co-occurring disorders
may share a common etiology and be different phenotypic expressions of a common causal factor or factors.
They may also be linked by etiology or pathological
mechanism, but one disorder may be a milder version
of the other on a spectrum of severity of pathology or
impairment. One disorder may precede and increase
the risk for the occurrence of another disorder, making

Manifestations, Clinical Diagnosis, and Comorbidity

a person more vulnerable to developing the second


disorder. A second disorder may arise after a first as a
complication or residual phenomenon or scar. People with certain personality disorders and related Axis
I disorders may share common psychobiological substrates that regulate cognitive or affective processes or
impulse control. The Axis I disorders may be the direct symptomatic expression of dysfunctions in these
systems, whereas personality disorders may reflect
coping mechanisms and more general personality
predispositions arising from the same systems (Siever
and Davis 1991). This more comprehensive model of
disorder co-occurrence integrates aspects of the common cause, spectrum, and vulnerability hypotheses.
Axis I/Axis II co-occurrence may be viewed from the
perspectives of the course of a persons lifetime or the
current presenting illness. Lifetime rates will obviously
be higher. Patients with personality disorders who are
seeking treatment also tend to have elevated rates of
Axis I disorder co-occurrence, because the development
or exacerbation of an Axis I disorder is often the reason a
personality disorder patient comes for clinical attention
(Shea 1997). For disorder co-occurrence to be significant
from a scientific perspective, rates must be elevated
above those expected by chance, based on the rates of occurrence of the individual disorders in a given clinical
setting or population. From a treatment perspective, any
co-occurrence may be significant.
The personality disorders of Cluster Aparanoid,
schizoid, and schizotypalare linked by theory and
phenomenology to Axis I psychotic disorders such as delusional disorder, schizophreniform disorder, or schizophrenia. Few studies have actually documented these associations, however, possibly because of problems in
being able to differentiate between clinical presentations
of attenuated and full-blown psychotic symptoms that
warrant two diagnoses instead of just one. (This problem
in differential diagnosis is discussed later.) Oldham et al.
(1995) found elevated odds of a current psychotic disorder in patients with Cluster A personality disorders but
also found elevated odds for Clusters B and C personality disorders as well, suggesting less disorder specificity
than might be expected.
In contrast, Cluster B personality disorders, especially BPD, which is linked by theory and phenomenology to Axis I mood and impulse control disorders, have
repeatedly been shown to have high rates of co-occurring MDD and other mood disorders, substance use disorders, and bulimia nervosa (Oldham et al. 1995; Skodol
et al. 1993, 1999; Zanarini et al. 1989, 1998). Taking into
account co-occurrence expected by chance alone, however, neither Oldham et al. (1995) nor McGlashan et al.

71

(2000) substantiated the relationship between BPD and


MDD. In addition, several studies have shown significantly elevated rates of anxiety disorders, including
panic disorder and posttraumatic stress disorder, in patients with BPD (McGlashan et al. 2000; Skodol et al.
1995). ASPD is most strongly associated with substance
use disorders in clinical and general population samples
(Grant et al. 2004; Kessler et al. 1997; Morgenstern et al.
1997; see also Chapter 30, Substance Abuse). This association supports an underlying dimension of impulsivity
or externalization (acting-out and being at odds with
mainstream goals and values) shared by these disorders
(Krueger et al. 1998, 2002).
Cluster C personality disorders, especially avoidant
and dependent personality disorders, are linked by
theory and phenomenology to anxiety disorders (Tyrer
et al. 1997). A number of studies have demonstrated
high rates of co-occurrence of avoidant personality disorder with MDD, agoraphobia, social phobia, and obsessive-compulsive disorder (Herbert et al. 1992; Oldham et al. 1995; Skodol et al. 1995). The co-occurrence
rates between avoidant personality disorder and social
phobia (particularly the generalized type) have been so
high in some studies that investigators have argued
that they are the same disorder. Ways of deciding
whether two diagnoses are warranted are discussed below under Problems in Differential Diagnosis. Several studies have indicated that dependent personality
disorder co-occurs with a wide variety of Axis I disorders, consistent with the notion of excessive dependency as a nonspecific maladaptive behavior pattern
that may result from coping with other chronic mental
disorders (Skodol et al. 1996). OCPD may be specifically linked to obsessive-compulsive disorder; however, an association between them has only inconsistently been found.
Paying attention to the co-occurrence of Axis I
and Axis II disorders is more than an intellectual exercise. The presence of an Axis I disorder in a patient
with a personality disorder may suggest a more specific treatment approach, either with pharmacological agents, psychotherapy, or self-help groups (as in
the case of substance use disorders), that will favorably affect outcome in these patients. Conversely, the
presence of personality disorder in a patient with an
Axis I disorder often indicates greater and more
widespread levels of impairment (Jackson and Burgess 2002; Skodol et al. 2002), more chronicity (Grilo
et al. 2005; Hart et al. 2001), and an overall poorer response to treatment requiring more intensive and
prolonged care (Reich and Vasile 1993; Shea et al.
1992).

72

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Co-Occurrence of Personality Disorders With


Other Personality Disorders
When thorough assessments of the full range of Axis II
disorders are conducted, as in research studies employing semistructured interviews, approximately half of
patients receive more than one personality disorder diagnosis. Patterns of co-occurrence of personality disorders generally follow the DSM cluster structure (i.e,
schizotypal personality disorder occurs more frequently
with paranoid and schizoid personality disorders than
with personality disorders outside Cluster A). These
patterns are consistent with factor-analytic studies that
support the clustering of personality disorders in DSM
(Kass et al. 1985; Sanislow et al. 2002). Some personality
disorders, however, particularly those in Cluster C,
show associations with personality disorders from other
clusters. Dependent personality disorder commonly occurs in patients with BPD, which makes clinical sense
because patients with BPD can display regressive, clinging, and dependent behavior in interpersonal relationships. Some personality disorders rarely co-occur.
OCPD and ASPD would be an exceedingly rare combination, because the careful planning and work orientation of OCPD are the antithesis of the impulsivity and
irresponsibility of ASPD.

Multiple Overlapping Personality Syndromes


Elevated rates of personality disorder co-occurrence
raise questions about the appropriate application of
DSM-IV-TR categories to phenomenology that rarely appears to have discrete boundaries. Although DSM-IV-TR
clearly stipulates that for many patients, personality disturbance would frequently meet criteria for more than
one disorder, clinicians have found the practice of diagnosing multiple disorders conceptually difficult and
therefore seldom attempt such diagnoses.
Prior to DSM-III-R (American Psychiatric Association 1987), part of the problem had been that most of
the personality disorders were defined as classical categories (Cantor et al. 1980)that is, ones in which all
members clearly share certain identifying features.
Classical categories imply a clear demarcation between members and nonmembers, but natural phenomena rarely fit neatly into such categories.

CATEGORIES VERSUS DIMENSIONS


OF PERSONALITY
Traditionally, in much of the psychological literature,
personality has been described and measured along

certain dimensions (Frances 1982). Dimensions of personality frequently are continuous with opposite traits
at either end of a spectrum, such as dominant-submissive or hostile-friendly. People can then vary in the extent to which each of the traits describes them. Dimensional models of personality diagnosis appear to be
more flexible and specific than categorical models
when the phenomenology lacks clear-cut boundaries
between normal and abnormal and between different
constellations of maladaptive traits, as seems true of
personality disturbance (Widiger et al. 1987). Scaled
rating systems have been devised to transform Axis II
disorders into dimensions (Kass et al. 1985; Oldham
and Skodol 2000), but they are not representative of
dimensional approaches currently in wide use. Dimensional models of personality disorders are being
seriously considered for DSM-V. They are discussed
in detail in Chapter 3, Categorical and Dimensional
Models of Personality Disorders.

CLASSICAL VERSUS PROTOTYPAL


CATEGORIES
Prototypal models have been shown to be more accurate than classical models in categorizing various natural phenomena. In the prototypal model, no defining
feature is considered to be absolutely necessary, nor is
any combination of features sufficient. Membership is
heterogeneous, and boundaries overlap. There are
few, if any, pathognomonic signs. The diagnostic criteria for a prototypal model are polythetic rather than
monothetic. Monothetic classifications are those in
which categories differ by at least one feature that is
shared by each of its members. In contrast, in polythetic
classifications, members share a large proportion of features but do not necessarily share any particular feature (Widiger and Frances 1985). In the prototypal
model, polythetic criteria would vary in their diagnostic value, and members would differ in terms of their
prototypicality.
A prototypal approach to personality disorder
classification is conceptually satisfying because of its
flexibility, the inherent heterogeneity of the categories,
and the acceptance of overlapping boundaries and
many borderline cases. From a conceptual point of
view, some of the diagnostic problems alluded to earlier would be lessened with a prototypal approach; for
example, multiple diagnoses and variability within diagnostic groups would be expected.
Monothetic categories are inherently more difficult
to recognize or diagnose because disagreement on any

Manifestations, Clinical Diagnosis, and Comorbidity

one of the required defining features results in disagreement on the diagnosis. With polythetic criteria,
because no single symptom is required for a diagnosis,
clinicians can disagree about an individual symptom
and still agree on the diagnosis, provided the particular symptom was not the one that met the minimum
threshold for the number of symptoms required for
the diagnosis.
DSM-IV-TR has a prototypal model for all personality disorders, defined by polythetic criteria sets. The
number of features listed varies from seven to nine,
with cut-points for the diagnosis at four or five required symptoms. ASPD is an exception in that it is
still a mixed category. A current age of 18, a childhood history of conduct disturbance, and irresponsible and antisocial behavior as an adult are necessary
for an ASPD diagnosis.
All DSM-IV-TR criteria carry equal weight; in a true
prototypal model, certain criteria would have more diagnostic significance. Research studies have demonstrated that for BPD, certain individual symptoms, such
as chronic feelings of emptiness and boredom (Widiger
et al. 1984) and suicidality or self-injury (Grilo et al.
2001, 2004a), have a higher value in predicting a diagnosis than other symptoms, such as impulsivity. Similar
highly predictive individual symptoms have been suggested for schizotypal personality disorder (e.g., odd
behavior, odd thinking or speech, constricted affect)
and OCPD (e.g., miserliness, preoccupation with details and rules) (Grilo et al. 2001). Predictive symptoms
need to be determined for all of the personality disorders and need an appropriate weighting system devised for them.
The currently required numbers of symptoms for
each of the personality disorders are arbitrary. Arguments have been made that fixed cut-points for diagnosis are inappropriate and inefficient. Appropriate
cut-points are actually dependent on the base rate of
the syndromethat is, how common it is in the population. For a particular symptom to be more likely to
indicate the presence of a syndrome rather than its absence, the ratio of the base rate to one minus the base
rate must exceed the ratio of the false-positive rate to
the true-positive rate (Finn 1982). If a symptom correctly identifies 80% of patients with the disorder and
misidentifies only 25% without the disorder, then at
least 24% of the patients must have the disorder or the
symptom will misclassify more patients than it correctly classifies. Therefore, if the disorder occurred
less often, given the presence of any one symptom
with the above diagnostic value, it would be more efficient never to diagnose the disorder because the clinician would then be wrong less often!

73

As the base rate of a syndrome changes, the efficiency of any cut-point also changes. If the base rate is
high, it is more efficient to move the cut-point down
because, with a high base rate, there is less chance of
missing the diagnosis and more chance of correctly
identifying the cases. If the base rate is low, the cutpoint should be increased, because it is increasingly
likely to incorrectly identify a noncase as a case. A
higher threshold for the symptoms would guard
against this error. Finally, the relative costs and gains
of correctly or incorrectly diagnosing cases could be
factored into establishing cut-points. This depends on
how the diagnosis is used or the implications of a
missed diagnosis and is referred to as the utility.
Studies need to be done to determine cut-points for
the various personality disorders that would be optimal in a variety of clinical settings and that might take
into account the utilities of the diagnostic decisions.
Some personality disorder researchers advocate a
prototype matching approach to the diagnosis of personality disorders rather than the current DSM procedure, which continues to involve making present/
absent judgments about individual criteria (Shedler
and Westen 2004). They would replace the diagnostic
criteria sets with descriptions of various personality
disorder prototypes in paragraph form and ask clinicians to rate the degree of similarity between the prototypes and the patient undergoing evaluation. They
argue that a prototype matching approach allows the
clinician to consider individual criteria in the context
of the whole personality disorder description, such
that no single criterion can make or break the diagnosis. They also argue that a prototype matching approach is closer to the way clinicians make personality
disorder diagnoses in actual practice.

PROBLEMS IN DIFFERENTIAL DIAGNOSIS


In this section, the individual personality disorders are
grouped according to the three descriptive clusters in
DSM-IV-TR: 1) the odd or eccentric, 2) the dramatic,
emotional, or erratic, and 3) the anxious or fearful. Although these clusters were originally introduced solely
to emphasize the descriptive similarities among the
disorders grouped together, some empirical evidence
has shown the validity of the clusters (Kass et al. 1985;
Sanislow et al. 2002; Widiger et al. 1987).

Odd or Eccentric Cluster


Paranoid, schizoid, and schizotypal personality disorders constitute the odd or eccentric cluster. Disorders

74

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

in this cluster share beliefs that are associated with


traits of social awkwardness, being ill at ease in social
situations, and social withdrawal.

Paranoid Personality Disorder


People with paranoid personality disorder are characterized by pervasive distrust and suspiciousness of others. Because of their expectation that others will exploit,
harm, or deceive them in some way, they are reluctant
to confide in others and therefore may seem distant and
removed. This type of social discomfort or withdrawal
is distinguishable from that of the schizoid patient because the schizoid patient appears not to care, whereas
the paranoid patient cares a great deal. People with
paranoid personality disorder are therefore the opposite of those with schizoid personality disorder in their
responses to the praise or criticism of others. Whereas
schizoid people are indifferent, paranoid people are extremely sensitive, very easily slighted, quick to take offense, ready to counterattack, and prone to bear grudges.
Patients with paranoid personality disorder can be distinguished from those with schizotypal personality disorder by the absence of symptoms such as magical thinking, unusual perceptions, and odd speech.
Patients with BPD also may react angrily to seemingly minor provocations, but they are not generally
suspicious and distrustful. Patients with narcissistic
personality disorder may appear distant from others,
particularly when they perceive threats to their selfesteem, but not because of general distrust. Patients
with avoidant personality disorder also are reluctant to
confide in others, but this reticence is based on their insecurity and not because they fear exploitation or harm.
Another point relevant to the differential diagnosis
of paranoid personality disorder is the relationship of
nonpsychotic suspiciousness and ideas of reference to
the delusions characteristic of a delusional disorder or
paranoid schizophrenia. The distinction rests on the
degree to which reality testing is impaired. In brief, in
paranoid personality disorder, the person can at least
entertain the possibility that his or her suspicions are
unfounded or that he or she is overreacting. Also, the
perceived threats of the person with a paranoid personality disorder are more likely to come from known
other people in the environmenta neighbor or a
co-worker, for instanceor from common institutions
such as the government or the utility company rather
than from bizarre sources. In cases in which beliefs of
expected harm or persecution are firmly held and result in extensive effects on behavior, paranoid personality disorder is not a sufficient diagnosis: the diagnosis of a psychotic disorder is warranted.

Schizoid Personality Disorder


There is some question of the validity of schizoid personality disorder as a distinct personality disorder. People who would have received the diagnosis of schizoid
personality traditionally might be diagnosed as either
schizoid, schizotypal, or avoidant by DSM-IV-TR criteria. In the few studies looking at the full range of personality disorders (e.g., Oldham et al. 1995; Pfohl et al.
1986), schizoid personality disorder was uncommon. It
must be remembered, however, that subjects in clinical
studies are selected by virtue of their seeking treatment;
schizoid people, by their very nature, are less likely to
seek treatment because subjective distress about their
attitudes and behavior is apt to be low, and impairment
would be evident only in the eyes of others, whom they
typically avoid. The crucial distinguishing features of
schizoid personality are that the person is detached
from social relationships and has a restricted range of
emotions in interpersonal settings. Although all Cluster
A personality disorders are characterized by social isolation, schizoid personality disorder can be distinguished from paranoid personality disorder by a lack
of general suspiciousness and from schizotypal personality disorder by a lack of cognitive and perceptual distortions. The more passive detachment and limited desire for social intimacy serves to distinguish schizoid
persons from avoidant personswho are also socially
isolated because they are petrified by their fear of rejection, despite a great desire for relationships (Trull et al.
1987). Patients with OCPD are often interpersonally
constrictedbut this is because they use excessive devotion to their work to protect themselves from their
discomfort with the emotions that arise in intimate relationships. Schizoid personality disorder is distinguished from psychotic disorders by the absence of delusions and hallucinations.

Schizotypal Personality Disorder


Schizotypal personality disorder was first introduced
in DSM-III. The criteria for schizotypal personality
disorder were developed in a study conducted by
Spitzer et al. (1979). The criteria were developed from
the case records of the borderline schizophrenic relatives of people genetically related to probands with
schizophrenia in the Danish Adoption Studies of
Schizophrenia (Kety 1983). They were intended to
help clarify the murky diagnostic area of borderline
patients.
The key defining features of schizotypal personality disorder are the soft, nonpsychotic symptoms that

Manifestations, Clinical Diagnosis, and Comorbidity

resemble those seen in more florid form in schizophrenia and make schizotypal patients appear eccentric.
These include magical thinking, ideas of reference, recurrent illusions, odd speech, and paranoid ideation.
Among the problems in differential diagnosis are how
to distinguish these features from their psychotic
counterparts and how to distinguish schizotypal patients from others in the odd, eccentric cluster.
The distinction between the soft, suggestive signs
and the full-blown psychotic symptoms rests on the
conviction regarding the beliefs, the vividness of the
illusions, and the degree of disorganization of speech.
Illusions are misperceptions of real external stimuli
and are thus distinct from hallucinations, in which a
sensory perception occurs without external stimulation of the sense organs. An example of a visual illusion might be mistaking a shadow for a real person or
seeing ones face change in a mirror. An auditory illusion might be hearing derogatory remarks made in
muffled conversation heard from a distance. In the
case of an illusion, the person can usually consider the
possibility that his or her perception was mistaken.
Odd speech may be tangential, circumstantial,
stilted, vague, or overly metaphorical. It differs from
loosening of associations in that it is generally more
understandable, although coherence is obviously
along a continuum. If a person with schizotypal personality disorder develops full-blown delusions or
hallucinations, then the diagnosis becomes schizophrenia because the premorbid symptoms of schizotypal personality disorder almost invariably would
meet the 6-month duration requirement for schizophrenia as prodromal symptoms.
The likelihood of schizotypal personality disorders evolving into schizophrenia is not fully established. What is known about the historical forerunners
of the diagnosis of schizotypal personality disorder
simple and latent schizophreniasuggests that only a
limited proportion actually develops schizophrenia
on follow-up. The only long-term follow-up study of
DSM-IIIdefined schizotypal personality disorder
was conducted by McGlashan (1986). He found that
pure schizotypal personality disorder had a better
prognosis than schizophrenia but worse than BPD.
The frequency with which schizotypal personality disorder became schizophrenia was 17% in the 15 years
of follow-up (Fenton and McGlashan 1989). If a patient with a past history of schizophrenia currently
displays symptoms of schizotypal personality disorder, the symptoms are usually referred to instead as
residual schizophrenia.
The schizoid/schizotypal distinction is made on
the basis of the presence of the psychotic-like symp-

75

toms in the latter. Schizotypal patients are more odd


and eccentric than patients with paranoid personality
disorder and have perceptual as well as cognitive distortions. Patients with BPD may have transient paranoid and dissociative symptoms accompanied by
strong affects, such as anger or anxiety, in response to
the stress of perceived abandonment. Although the
psychotic-like symptoms of patients with schizotypal
personality disorder may also worsen with stress, this
response is less likely to occur in the context of a disruption in an interpersonal relationship and to be accompanied by strong affect. Patients with BPD periodically withdraw from social relationships in the face of
disappointment, whereas patients with schizotypal
personality disorder more generally avoid social involvement and are not typically impulsive.
The following vignette illustrates the case of a socially isolated person that raises differential diagnostic
questions.

Case Example
A videotaped interview of a 30-year-old bachelor
was shown to 133 American and 194 British psychiatrists in the late 1960s as part of the United States
United Kingdom comparative study of psychiatric diagnosis (see Skodol 1989).
Problems began for the patient when he was 13 or 14
years old. He described himself as insecure and very
dependent on his mother for emotional support. Although he claimed he sometimes did well in high
schoolplayed football, boxed, acted, and played
the trumpetat other times, he said, he was afraid to
go to school and would stay home with his mother.
He said he was afraid other kids would pick on him
and he would get into a fight. He attended several
colleges but did not study and accumulated only
1 years of credit.
He then joined the army but lasted only 5 months.
He was hospitalized briefly, at age 19, at Walter Reed
Hospital but claims he was told that there was nothing wrong. He states that he felt like a little boy and
wanted to go home to his mother. He said he broke
down and screamed and cried like a baby.
His most recent hospitalization was his fifth. The
longest had been for 5 months; the others, for several
days to several weeks. In all cases and on other occasions he requested hospitalization. He was often refused and told that he did not need hospitalization
but should go to work. He had been treated with a
variety of medicines, including phenothiazines, and
had received 20 electroconvulsive treatments as an
outpatient.
Other problems he describes were periodic abuse
of drugs, including alcohol, barbiturates, opioids, and
amphetamines. He reports periods of not being able
to get out of bed, shave, or shower; he denies de-

76

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

pressed mood or symptoms of a depressive syndrome. He also denies grandiosity or other symptoms
of a manic syndrome. He has worked very sporadically and states that he purposely fails at tasks. He
says he makes friends but quickly loses them. He has
not seen any friend for the past 6 months.
On the videotaped interview, the patient has just
described an incident in which he developed a paralyzed arm, which his psychiatrist called a hysterical symptom.
Interviewer: What other sorts of things have
happened to your body?
Patient: Well, one thing is that no matter how
I look to you now, my facial appearance
changes sometimes, unbelievably. Now,
a lot of doctors thought I was exaggerating, but my own mother says its true.
Sometimes my face just blows up, my
nose gets wider, my eyes close up, and
(giving his cheek a twist), I cant feel
nothinlike this.
Interviewer: What does this to you?
Patient: Simply, if it didnt, Id have no reason to tell myself that Im afraid to go
out into the world.
Interviewer: You mean that your face actually
does swell up, or that you imagine it?
Patient: It actually does! I swear on my heart
that I never imagined anything, or seen
anything that wasnt there.
Interviewer: How long has this been happening to you?
Patient: Ten years.
Interviewer: What happens if you look in
the mirror?
Patient: I dont.
Interviewer: Why not?
Patient: To avoid it. I try to forget about it.
I know that my basic problem isnt my
faceI used to think it was. Now I know
itll change when the basic problem goes
away.
Interviewer: Does it frighten you that this
happens?
Patient: It used to. I used to think that I was the
owner of a fantastic symptom that was
totally unbelievable, plus I couldnt get
any medical man to believe me. Finally,
I went to one or two psychiatrists who
told me theyd seen it before, maybe not
the face, but a physical change can take
place.
Interviewer: If you go out in public, do you
feel self-conscious about this?
Patient: Thats whats amazing. When Im
sick like this I dont feel self-conscious.
I could be as ugly as the day is long. But
when Im well, and look my best, or get
attention from people, I cant stand it.
Interviewer: What do you do then?

Patient: I withdrawinto myself. This way


nobody is going to come up to me.
I wont be forced to reactHello; goodbye. Converse. Talk. Walk. Work.
Interviewer: I see you wear dark glasses.
Patient: Yeah, well in the safety of my own
house I feel OK, but if I walk out onto
the street, it hits me: Where? How?
Who do I go to? Theres 30 billion people. Who do I speak to? Where do I go?
Next thing I know, Im paranoid.
Interviewer: What do you mean, paranoid?
Patient: People look at me. They could be
saying anything. Hes good-looking
or Hes ugly. But all I feel is Oh, my
God! I cant stand this! People looking
at me! You know, when I get looked at
because I look terrible, that doesnt
frighten me. But should I feel good and
get some attention, you know, I get sick.

The patient depicted in this vignette was fascinating


because there was more disagreement between American and British psychiatrists on the appropriate diagnosis than on any other case in the study (Kendell et al.
1971). Sixty-nine percent of American psychiatrists in
the late 1960s diagnosed this man as having schizophrenia; only 2% of British psychiatrists did so. The
most common British diagnosis was personality disorder, usually hysterical. The next most common diagnosis by British psychiatrists was neurosis. Most mental
health clinicians in the United States to whom I have
presented the videotape corresponding to this vignette
agree that on Axis I, diagnoses of mixed substance
abuse and conversion disorder are warranted. A factitious disorder is the second most frequently chosen diagnosis. On Axis II, using DSM-III criteria, most clinicians chose schizotypal personality disorder with
histrionic features. With the expansion of the concept of
avoidant personality disorder in DSM-IV to include
more prominent fearfulness, I suspect that clinicians using DSM-IV would also note avoidant features.

Dramatic, Emotional, or Erratic Cluster


The dramatic, emotional, or erratic cluster includes
antisocial, borderline, histrionic, and narcissistic personality disorders. These highly overlapping disorders share the characteristics of reactive emotionality
and poor impulse control.

Antisocial Personality Disorder


ASPD is unique among personality disorders in that it
can be reliably diagnosed, even in clinical settings. It is
less difficult to recognize because its characteristic
pattern of behaviors, which disregard or violate the

Manifestations, Clinical Diagnosis, and Comorbidity

rights of others, beginning in adolescence, are identified by very explicit lists of antisocial activities. DSMdefined ASPD has also been widely criticized, however, for an overemphasis on overt criminal acts at the
expense of the personality traits of psychopathy, such
that it is overdiagnosed in criminal or forensic settings
and underdiagnosed in noncriminal settings (Widiger
and Corbitt 1996).
Patients with narcissistic personality disorder
share some of the arrogant, exploitative, nonempathic
characteristics of patients with ASPD but usually are
not impulsive or physically aggressive, nor do they
have a history of childhood conduct disorder. Patients
with narcissistic personality disorder who engage in
criminal behavior are most likely to commit whitecollar crimes. Patients with histrionic and borderline
personality disorders may be impulsive and manipulative but are seeking attention and nurturance, respectively, rather than profit, power, or material gain.
Patients with BPD may be overrepresented in criminal
populations, especially among women (see Chapter
36, Correctional Populations: Criminal Careers and
Recidivism). If patients with paranoid personality
disorder engage in antisocial behavior, it is based on a
desire for revenge over a perceived slight, rather than
for personal gain or exploitation of others.
Conduct disorder is a diagnosis for a repetitive and
persistent pattern of behavior among children or adolescents under 18 years of age in which the rights of
others or societal norms are violated. The restriction of
ASPD to persons over 18 means that the pattern has to
have persisted into adult life, because many childhood
conduct problems may remit or may lead to other
mental disorders.
Other mental disorders such as psychotic disorders and mood disorders can lead to breaking of laws
and antisocial acts. Schizophrenic or manic episodes
preempt the diagnosis of ASPD. Patients with substance-related disorders (see Chapter 30, Substance
Abuse) may engage in antisocial behaviors such as illegal drug selling or theft to obtain money for drugs.
Both diagnoses may be given, even if some of the criteria met for ASPD are related to drug use. When antisocial behavior occurs that is not a part of the full pattern of ASPD or is not due to another mental disorder
such as schizophrenia, then the V code category of
adult antisocial behavior is appropriate.

Borderline Personality Disorder


BPD has generated by far the most extensive and intensive research of all of the DSM-IV-TR personality
disorders. This research interest simply reflects the in-

77

tense clinical interest in borderline patients, who seem


to have swelled the ranks of inpatient hospitals and
outpatient practices of the past 35 years. The two major questions that have been asked are 1) What are the
borders of borderline? and 2) What are the key clinical features of this disorder?
The criteria for BPD were originally defined by
Spitzer et al. (1979) in an effort to delineate which patients clinicians referred to as borderline. These investigators found two overlapping sets of descriptive
items, a set reflecting instability of affect, identity, and
impulse control and another reflecting eccentricity of
thought, speech, and behavior. The former became the
criteria for BPD, and the latter for schizotypal personality disorder, in DSM-III.
Although traditionally, and in psychoanalytic
terms, borderline patients were thought to occupy a
border between psychosis or schizophrenia and
neurotic disorders, evidence accumulated, based on
the validation techniques of family history, treatment
response, and outcome on follow-up, that indicated
that BPD bore much more of a relationship to affective
disorders than to schizophrenia (e.g., Akiskal et al.
1985; Snyder et al. 1982). This led many clinicians (and
researchers) into the diagnostic dilemma of attempting to distinguish whether a particular patient has
BPD or an affective disorder.
This dilemma is a product of asking the wrong
question. The appropriate question is which patients
with BPD also have a mood disorder. The relevancy of
this question for clinical practice is supported by the
most recent reviews of this area of differential diagnosis. Gunderson and colleagues (Gunderson and Elliot
1985; Gunderson and Phillips 1991) examined four hypotheses about the interface between BPD and affective disorder: 1) that affective disorder is primary and
that borderline character traits such as drug use and
sexual promiscuity arise in an attempt to alleviate
depression; 2) that BPD leads to affective disorder
(depression) as a result of problems that result from
primary deficits in impulse control, maintaining stable
interpersonal relationships, and sense of self-esteem;
3) that the two disorders are independent, but because
both occur frequently in the population, they are often
seen together; and 4) that they are related, but in a
nonspecific fashion. The data, the authors argued,
supported none of the hypotheses as stated. They
were most consistent with the independence hypothesis, but the two disorders co-occurred more frequently
than would be expected by chance.
Recently, Gunderson et al. (2004) have reexamined
the relationship of BPD and MDD from a longitudinal

78

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

perspective. They found that although the courses of


BPD and MDD could be independent, improvements
in MDD were more likely to occur following improvements in BPD than the reverse. These results support
the view that BPD is a fundamental form of psychopathology that accounts for co-occurring depressions
and that these depressions should be understood as
epiphenomena of the abnormal sensitivity and interpersonal disappointments of patients with BPD. This
view is further supported by the qualitative differences
in the depressive experiences (e.g., the marked reactivity of mood to identifiable events [Gunderson 1996]) of
patients with BPD compared with those with MDD
(Rogers et al. 1995; Westen et al. 1992); by follow-up
studies that fail to show that BPD evolves into more
typical mood disorders over time (Grilo et al. 1998);
and by the relatively modest response of patients with
BPD to antidepressant medications (Koenigsberg et al.
1999; Soloff et al. 1998).
From a clinical perspective, the important distinctions to be made, therefore, are among BPD alone, BPD
in association with a mood disorder in the depressive or
bipolar spectra, and affective disorder alone. These distinctions are facilitated by the DSM-IV-TR multiaxial
system because Axes I and II are considered separately,
and multiple diagnoses can be listed on each axis.

Case Example
The following case example is adapted from Skodol
(1989).
A 37-year-old single woman, a bookkeeper for a
building restoration and waterproofing company,
was evaluated for hospital treatment. She described
herself as chronically and severely depressed since
the age of 18 and bulimic since her early 20s. She
said, Ive cried every day for the past 10 years. She
had an extremely low opinion of herself: You have
never met anyone as bad as I am, I guarantee it."
She had had 14 years of therapy with a halfdozen therapists. She typically became very attached
to them, then reacted extremely negatively, sooner
or later, when they let her down. Once, when a therapist would not allow her to extend a session beyond
her time, she picked up an ashtray and threw it at
him. Another time, she waited for one of her therapists after his day was over, lay down in front of his
car, and would not let him go home before he talked
more to her. On still another occasion when she was
angry at a therapist, she took a razor blade from her
purse and cut her wrist in the therapists office.
Many medications had been tried for both the
depression and the bulimia. She had been on Librium and Valium many years before, then Elavil, Tofranil, Mellaril, and lithium, followed by Xanax, Par-

nate, and Nardil; most recently she had been given


Prozac, Zoloft, and Effexor. Occasionally, the depression abated slightly for maybe 1 week. As far as
her concern with her weight and her binge eating,
she claimed nothing helped. Her weight had ranged
from a low of 110 to a high of 130. She claimed that
she had taken up to 70 laxatives in a week and had
vomited every day for almost 10 years. She also had
panic attacks and had abused alcohol, cannabis, and
stimulants in the past.
The patient continued to work, although she did
not get along well with her coworkers. I know people dont like me. Im just a lazy, nasty person. Some
of them probably think Im grotesque. Im sure
theyre also laughing at me. Who wouldnt? Im an
absurdity. The patient had not had a date in 8 years
and had only a few female acquaintances.
A research interview indicated that the patient
met DSM-IV-TR criteria for five (!) personality disorders: avoidant, obsessive-compulsive, schizotypal,
histrionic, and borderline. The BPD was rated severe.

Standard treatments for major depression (or


bulimia) are no match for this womans personality
psychopathology. It is not difficult to conceptualize her
overall maladjustment as being so severe that minor
improvements in mood would be insignificant to her
or even unacceptable, given her self-defeating tendencies. A skeptical clinician might argue that given the
patients tendencies to exaggerate, manipulate, and
provoke, it would not be possible to accurately assess
the state of her mood in response to treatment. This
problem raises the question of which components of a
mood disorder are most likely to be affected by Axis II
psychopathology. Clearly, in work with patients with
severe personality disorders, the subjective state of the
patient is very resistant to change. Improvement may
be evident only by objective criteria, from the perspective of either the clinician or of a significant other in the
persons life.
Other Axis I disorders, such as anxiety disorders,
substance-related disorders, eating disorders, somatoform disorders, dissociative disorders, and psychotic
disorders, may also complicate the course of BPD (Zanarini et al. 1998; Zimmerman and Mattia 1999a). Cooccurrence of BPD with substance-related and eating
disorders suggests that BPD lies on a spectrum of disorders of impulse control (Siever and Davis 1991). A new
criterion in DSM-IV for transient, stress-related paranoid ideation or severe dissociative symptoms may
raise new issues in differentiating dissociative and psychotic disorders from BPD (the reactive, stress-related
nature of the symptoms characterize BPD [Sternbach et
al. 1992]). Again, however, in these instances the clinician should not necessarily pose the differential diagnosis in terms of either/or but instead as both/and.

Manifestations, Clinical Diagnosis, and Comorbidity

BPD overlaps extensively with histrionic, narcissistic, antisocial, and dependent personality disorders.
Patients with histrionic personality disorder can be
manipulative and experience rapidly shifting emotions but are not self-destructive, angry, or empty as
are patients with BPD. Patients with narcissistic personality disorder often react angrily to provocation
but have more stable identities and lack the problems
of impulse control, self-destructiveness, and fears of
abandonment seen in BPD. Patients with ASPD are
manipulative for personal gain, whereas those with
BPD are manipulative in order to get their needs met.
Both borderline and dependent personality disorders
are characterized by fears of losing the support of caretakers, but patients with BPD react to threats of loss of
such a person with angry demands, whereas the patient with dependent personality disorder becomes
more acquiescent and submissive.

Histrionic Personality Disorder


Histrionic personality disorder is defined in DSM-IVTR by excessive emotionality and attention-seeking behavior. In clinical and research settings, the features of
histrionic personality disorder overlap considerably
with those of other disorders in this cluster, especially
the narcissistic and borderline, and with dependent
personality disorder. Although histrionic patients may
make up a large proportion of psychotherapy patients,
they have not been well studied in terms of DSM-IVTR criteria.
The diagnostic overlap of histrionic with narcissistic personality disorder is possible because of the traits
and behaviors that the two have in common. Histrionic personality disorder includes incessant drawing
of attention to oneself and egocentrism; narcissistic
personality disorder includes a grandiose sense of
self-importance, entitlement, interpersonal exploitiveness, and lack of empathy. Patients with narcissistic
personality disorder usually want recognition because
of their superiority, whereas patients with histrionic
personality disorder will allow themselves to be
viewed as weak and dependent if doing so attracts
attention. When criteria for both disorders are met,
both diagnoses should be given.
Patients with BPD are frequently histrionic. Histrionic patients are demanding and manipulative. BPD
patients display inappropriate, intense anger, perform
physically self-damaging acts, and are demanding
and manipulate others. Histrionic patients lack the
more malignant characteristics of BPD. These patients,
referred to in the classic literature as hysterical, may be
very vain and self-indulgent, always drawing atten-

79

tion to themselves or craving action and excitement,


without having angry outbursts, making suicidal
threats or gestures, or feeling empty.
Another problem in making a diagnosis of histrionic personality disorder is that the symptoms are difficult for the patient to recognize. A patient who overreacts to minor events in most cases does not consider
the reaction excessive or the event minor. Few patients
are aware that others consider them shallow or manipulative or that their speech is overly impressionistic.
Therefore, histrionic personality disorder is a diagnosis that often requires the input of third-party informants. Fortunately, histrionic traits are usually displayed to the therapist, and observation is therefore of
great diagnostic value.
Patients with histrionic personality disorder may
be especially prone to Axis I disorders in the somatoform disorders class. The clinician should therefore be
alert to the possible additional diagnoses of somatization disorder, conversion disorder, pain disorder, hypochondriasis, or body dysmorphic disorder.

Narcissistic Personality Disorder


The hallmark features of narcissistic personality disorder in DSM-IV-TR are a grandiose sense of self-importance or uniqueness, preoccupation with fantasies of
success, an excessive need for admiration, and interpersonal relationship problems, such as feeling entitled, exploiting others for personal gain, and failing to
empathize with the feelings of others.
Overlap with other disorders in Cluster B has been
described previously. Both patients with narcissistic
and with obsessive-compulsive personality disorders
may appear perfectionistic, but patients with OCPD
are self-critical, whereas those with narcissistic are not.
Grandiosity is a symptom of a manic or hypomanic
episode, but the absence of an abrupt onset of elevated
mood and impairment in functioning help to distinguish narcissistic personality disorder from bipolar
disorders. Chronic use of certain substances, such as
cocaine, can also lead to grandiose, self-preoccupied
behavior patterns.
The diagnosis of narcissistic personality disorder
presents the difficult problem of translating concepts
of psychological functioning derived largely from the
psychoanalytic literature into descriptions of traits
and behaviors that can be recognized by clinicians
with diverse theoretical orientations. As Frances
(1980) has indicated, the psychoanalytic definition of
narcissistic personality disorder would include 1) deficits in object constancy, 2) incomplete internalization
and maturation of psychic structures and mechanisms

80

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

regulating self-esteem, and 3) immature grandiosity.


These problems are not easily recognized, especially
by nonanalytic clinicians, in one or two diagnostic interviews.
Deficits in object constancy are reflected in the
characteristic interpersonal disturbances of narcissistic personality disorder. Narcissistic people have an
inflated sense of their own self-importance and often
devalue the importance of others. Despite their outward air of superiority, self-esteem problems are evident when they react with disdain, rage, humiliation,
or emptiness in response to criticism or defeat. Immature grandiosity is reflected by narcissistic personality
disorder criteria describing a grandiose sense of selfimportance, preoccupations with fantasies of unlimited potential, beliefs in special and unique attributes,
and entitlement.

Anxious or Fearful Cluster


Avoidant, dependent, and obsessive-compulsive personality disorders make up the anxious and fearful
cluster. At least one factor analytic study (Kass et al.
1985) has shown that OCPD may not fit as well into
this group as the others.

Avoidant Personality Disorder


Avoidant personality disorder is characterized by social inhibition due to feelings of inadequacy and a fear
of being negatively evaluated by others. Both avoidant
and dependent personality disorders are characterized
by feelings of inadequacy, hypersensitivity to criticism,
and need for reassurance. In patients with avoidant
personality disorder, the concern is with avoiding embarrassment or humiliation; in patients with dependent
personality disorder, it is with being taken care of. The
two disorders often co-occur, however.
Items referring to exaggerating the difficulties or
risks of new but ordinary activities and situations, and
to embarrassment and social anxiety, make avoidant
personality disorder in DSM-IV-TR close in concept to
the phobic character style common in the psychoanalytic literature. Differentiating avoidant personality
disorder from social phobia, especially generalized
social phobia, can be difficult. Research has shown
that although there is significant co-occurrence of social phobia and avoidant personality disorder, they
are not synonymous, and patients can meet criteria for
one disorder without meeting criteria for the other
(Skodol et al. 1995). The concept of avoidant personality disorder is broader than that of generalized social
phobia in that it includes feelings of inadequacy, infe-

riority, and ineptness and a general reluctance to take


risks and engage in new activities.

Dependent Personality Disorder


Dependent personality disorder is characterized by
clinging and submissive behavior and an excessive
need to be taken care of. Dependent personality shares
with histrionic personality disorder a covariation with
gender, occurring more frequently in women (Kaplan
1983; Kass et al. 1983). It has been argued that this covariation results from a sex bias in the diagnostic criteria (Kaplan 1983), such that normal women conforming to their sex role stereotype will be labeled
abnormal because of a masculine bias about what constitutes healthy behavior.
One of the real problems in the diagnosis of dependent personality disorder is its threshold for clinical
significance. The earlier discussion in this chapter
about personality traits versus personality disorder is
germane. For dependent personality traits to indicate
a personality disorder, evidence of significant distress
or social or occupational impairment is necessary. If a
woman subordinates her needs to those of her husband to avoid losing him, then there would have to be
clear evidence that this behavior is damaging to her;
for example, if she does not choose other equally viable options for herself sociallyand with respect to
her family and living arrangementsbecause of her
inability to make her own decisions or act according to
her own needs. Another consideration is that a particular womans needs may be very different from her
husbands; she may desire greater affiliation and need
less self-determination in traditional areas of living
such as economic productivity. Keeping in mind the
need for strong evidence of the pathological nature of
the dependency may help guard against too many
false-positive diagnoses of women.
Many of the diagnostic criteria for dependent personality disorder resulted from a need to specify more
explicitly the kinds of dependent behaviors indicative
of the disorder and to emphasize their pathological
nature, for example, has difficulty making everyday
decisions without an excessive amount of advice and
reassurance from others and has difficulty initiating
projects or doing things on his or her own (American
Psychiatric Association 2000, p. 725). The person with
dependent personality disorder stays in poor relationships, goes along with others even when thinking they
are wrong, does demeaning things, and feels helpless
when alone all because of an inability to see himself or
herself as sufficiently competent. It is not the lack of
confidence per se that is significant for the person with

Manifestations, Clinical Diagnosis, and Comorbidity

dependent personality disorder but the pathological


use of relationships to attempt to deal with the perceived deficiency. Patients with dependent personality disorder are prone to having associated depressive
or adjustment disorders because they are so vulnerable to disappointments and disruptions in relationships.
Dependent personality disorder has been found to
co-occur with other personality disorders (Trull et al.
1987). The dependent-avoidant combination is particularly common.

Obsessive-Compulsive Personality Disorder


The essential features of OCPD are perfectionism, inflexibility, and control. OCPD does not overlap extensively with other personality disorders in this cluster.
OCPD shares with dependent and histrionic personality disorders the problem of being applied as a sex stereotypeonly this time referring to stereotypic male
behavior such as excessive devotion to work or insistence on getting ones way (Reich 1987). The same caution applies, therefore, for the clinician to document
the pathological nature of the behaviors and the impairment that results. This documentation is somewhat easier in the case of OCPD than dependent personality disorder because the disorder items in the
former, such as perfectionism, preoccupation with details, and excessive devotion to work, all explicitly
refer to how these traits interfere with functioning.
Perfectionism, for example, interferes with task completion, so that the patient is unable to complete a
project because his or her own overly strict standards
are not being met (American Psychiatric Association
2000, p. 729).
A significant distinction should be made between
OCPD and obsessive-compulsive (anxiety) disorder.
Patients with OCPD may not have true obsessions or
compulsionsthat is, recurrent, senseless thoughts or
repetitive, stereotypic behavior rituals. Occasionally,
the OCPD persons preoccupation with details, lists,
schedules, and the like may approach the threshold of
definition of obsessions or compulsions, but usually
these behaviors will feel ego-syntonic and purposeful to such a person.

Other Personality Disorder Types


Passive-Aggressive (Negativistic)
Personality Disorder
Passive-aggressive personality disorder is identified
by passive resistance to demands for adequate social
and occupational performance and by negative atti-

81

tudes. Passive-aggressive personality disorder is in


DSM-IV Appendix B, Criteria Sets and Axes Provided for Further Study. Long an official personality disorder in DSM, passive-aggressive personality
was placed in this appendix because it was not clear
whether the criteria identified a pervasive pattern of
thinking, feeling, and behaving characteristic of a personality disorder, or simply a single trait (i.e., resistance to external demands). Some attempt has been
made to emphasize cognitive and affective aspects of
the disorder. Thus, criteria refer to the persons believing that he or she is misunderstood and unappreciated by others, being critical or scornful of people in
authority, and becoming sullen and argumentative
(American Psychiatric Association 2000, p. 791).
The other major difficulty in the diagnosis of passive-aggressive personality disorder is that the behavior must be evident even in situations in which more
self-assertive behavior is possible. The military is usually given as the best example in which self-assertive
behavior is frequently not permitted and compliance
with the demands of others is required. Passive resistance to demands in this situation would not necessarily indicate a personality disorder. Sometimes it is
more difficult for the clinician to assess the rigidity of
the demands imposed by the external circumstances.
An example would be a job situation in which an employer indicated that there was much latitude for individual, independent initiative while subtly exerting
almost total control of the employees behavior.

Depressive Personality Disorder


Depressive personality disorder was a new disorder
introduced into DSM-IV Appendix B. This addition reflects an ongoing debate as to the appropriate characterization of chronic, mild depression as a personality
disorder or a mood disorder (Hirschfeld and Holzer
1994). Depressive personality disorder is manifested
by a pervasive pattern of depressive cognitions and behaviors, such as a gloomy and unhappy mood, beliefs
of inadequacy or worthlessness, critical and blaming
attitudes toward self and others, brooding, pessimism,
and guilt. The major problem in differential diagnosis
is in distinguishing depressive personality disorder
from dysthymic disorder. Studies have shown rates of
co-occurrence of these two disorders that vary widely,
from 18% to 95%, depending on the sample and the criteria used to make the diagnoses (e.g., Klein 1999;
Klein and Shih 1998; McDermut et al. 2003). Using
DSM-IV-TR criteria, depressive personality disorder
can be distinguished from dysthymic disorder by an
emphasis on cognitive, interpersonal, and intrapsychic
personality traits in the former and more physical,

82

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

vegetative symptoms, such as sleep and appetite


disturbance or fatigue, in the latter. When criteria for
both disorders are present, both diagnoses can be
made. Depressive personality disorder may also predispose to the development of episodes of major depressive disorder.

Self-Defeating Personality Disorder


The diagnosis of masochistic personality disorder was
by far the most frequently made diagnosis under the
rubric of other personality disorder in DSM-III. Masochistic personality disorder is thought by many clinicians to be a useful concept. In the process of revising
DSM-III, it quickly became a very controversial category, however, because feminist groups in particular
objected to what they viewed as its sexually discriminatory content. In part in response to these objections, the
diagnosis was renamed self-defeating personality disorder and was included in an appendix of DSM-III-R
as a proposed diagnostic category needing further
study. The category was dropped completely from
DSM-IV (American Psychiatric Association 1994) because its criteria described a behavior pattern common
to many other personality disorders and not sufficiently distinctive to represent a separate category
(Skodol et al. 1994), and data on its clinical utility and
external validity were sparse (Feister 1996).

Sadistic Personality Disorder


Some critics have also objected to the preoccupation of
mental health professionals with classifying the victim and ignoring the victimizer in situations in
which one person takes advantage of or abuses another. This mental set may be the result of victims being more likely than victimizers to seek help with emotional problems, but this likelihood does not justify
trying to understand the nature of only the victims
troubles.
Therefore, also included in the appendix of DSMIII-R were criteria for a new diagnosis that describes a
pattern of behavior characterized by cruel, demeaning,
and aggressive behavior for reasons other than sexual
arousal. This disorder was called sadistic personality disorder. The important points in the differential diagnosis
are to distinguish the behavior from those of the
paraphilias and from those of other disorders in the
differential diagnosis of violent behavior, such as
ASPD (see Chapter 31, Violence). Sadistic personality disorder was also dropped from DSM-IV because of
a paucity of empirical research to support its inclusion.

Personality Disorder Not Otherwise Specified


DSM-IV has a residual category for mixed or other
personality disorders. The mixed category is to be
used when a person with a personality disorder had
features of several of the specific personality disorder
types but does not meet the criteria for any one.
Other personality disorder is used when the clinician wants to diagnose a specific personality disorder
type that is not included in DSM-IV-TR (e.g., passiveaggressive, depressive, or self-defeating).
A common error in the use of the personality disorders section of DSM-IV-TR is assigning a diagnosis
of mixed personality disorder to a patient who meets
criteria for one disorder and has features of one or
more other personality disorders, or to a patient who
meets full criteria for more than one personality disorder. In the first instance, the clinician should diagnose,
for example, BPD with narcissistic and histrionic
traits; in the second instance, diagnoses of multiple individual personality disorders should be made.

SUMMARY
This chapter considers the manifestations, problems
in differential diagnosis, and patterns of comorbidity
of the DSM-IV personality disorders. Although considerable dissatisfaction has been expressed over the
DSM approach to these disorders and a major overhaul has been recommended by many researchers and
clinicians in the field (Clark et al. 1997; Shedler and
Westen 2004; Widiger 1991, 1993), the DSM approach
remains the official standard for diagnosing personality disorder psychopathology. Work on DSM-V has
recently begun, but its publication is not anticipated
until at least 2010. Therefore, even if a dimensional approach to personality disorders were to replace the
categorical approach in DSM-V, these changes would
not be implemented for several years.
Included in this chapter are descriptions of the
clinical characteristics of the 10 DSM-IV personality
disorders; discussions of problems in interviewing the
patient with a suspected personality disorder in state
versus trait discrimination, trait versus disorder distinctions, categorical versus alternative classificatory
approaches to personality disorder diagnosis, and diagnosis based on inferential judgments; and an overview of personality disorder comorbidity. Problems in
the diagnosis of each individual disorder are covered,
grouped according to the three DSM-IV clusters. Despite limitations in the DSM approach, personality

Manifestations, Clinical Diagnosis, and Comorbidity

disorders diagnosed by this system have been shown


in the past 25 years to have considerable clinical utility
in predicting functional impairment over and above
that associated with comorbid Axis I disorders, extensive and intensive utilization of treatment resources,
and in many cases, adverse outcomes.

REFERENCES
Adler DA, Drake RE, Teague GB: Clinicians practices in personality assessment: does gender influence the use of
DSM-III Axis II? Compr Psychiatry 31:125133, 1990
Akiskal HS, Chen SE, Davis GC, et al: Borderline: an adjective in search of a noun. J Clin Psychiatry 46:4148, 1985
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition. Washington,
DC, American Psychiatric Association, 1980
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition, Revised. Washington, DC, American Psychiatric Association, 1987
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition. Washington,
DC, American Psychiatric Association, 1994
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association,
2000
Andreoli A, Gressot G, Aapro N, et al: Personality disorders
as a predictor of outcome. J Personal Disord 3:307321,
1989
Beck AT, Freeman A, Davis DD, et al: Cognitive Therapy of
Personality Disorders, 2nd Edition. New York, Guilford, 2004
Benjamin LS: Interpersonal Diagnosis and Treatment of Personality Disorders, 2nd Edition. New York, Guilford,
1996
Bernstein DP, Cohen P, Skodol AE, et al: Childhood antecedents of adolescent personality disorders. Am J Psychiatry 153:907913, 1996
Blashfield RK, Herkov MJ: Investigating clinician adherence
to diagnosis by criteria: a replication of Morey and Ochoa
(1989). J Personal Disord 10:219228,1996
Bond MP, Vaillant JS: An empirical study of the relationship
between diagnosis and defense style. Arch Gen Psychiatry 43:285288, 1986
Cantor N, Smith EE, French RS, et al: Psychiatric diagnosis
as prototype categorization. J Abnorm Psychol 89:181
193, 1980
Carver CS, Scheier MF, Weintraub JK: Assessing coping
strategies: a theoretically based approach. J Pers Soc
Psychol 56:267283, 1989
Casey PR, Tyrer P: Personality disorder and psychiatric illness in general practice. Br J Psychiatry 156:261265,
1990

83

Clark LA, Livesley WJ, Morey L: Special feature: personality


disorder assessment: the challenge of construct validity.
J Personal Disord 11:205231, 1997
Dolan-Sewell RT, Krueger RF, Shea MT: Co-occurrence with
syndrome disorders, in Handbook of Personality Disorders: Theory, Research, and Treatment. Edited by Livesley WJ. New York, Guilford, 2001, pp 84104
Drake RE, Vaillant GE: A validity study of Axis II of DSMIII. Am J Psychiatry 142:553558, 1985
Feinstein AR: The pre-therapeutic classification of comorbidity in chronic disease. J Chronic Dis 23:455468, 1970
Feister SJ: Self-defeating personality disorder, in DSM-IV
Sourcebook, Vol 2. Edited by Widiger TA, Frances AJ,
Pincus HA, et al. Washington, DC, American Psychiatric Association, 1996, pp 833847
Fenton WS, McGlashan TH: Risk of schizophrenia in character
disordered patients. Am J Psychiatry 146:12801284, 1989
Finn SE: Base rates, utilities, and DSM-III: shortcomings of
fixed rule systems of psychodiagnosis. J Abnorm Psychol 91:294302, 1982
First M, Gibbon M, Spitzer RL, et al: Users Guide for the
Structured Clinical Interview for DSM-IV Axis II Personality Disorders. Washington, DC, American Psychiatric Press, 1997
Folkman S, Lazarus RS: The Ways of Coping Questionnaire.
Palo Alto, CA, Consulting Psychologists Press, 1988
Folkman S, Moskowitz JT: Coping: pitfalls and promise.
Ann Rev Psychol 55:745774, 2004
Frances A: The DSM-III personality disorders section: a commentary. Am J Psychiatry 137:10501054, 1980
Frances A: Categorical and dimensional systems of personality diagnosis: a comparison. Compr Psychiatry 23:516
527, 1982
Grant BF, Stinson FS, Dawson DA, et al: Co-occurrence of 12month alcohol and drug use disorders and personality
disorders in the United States: results from the National
Epidemiologic Survey on Alcohol and Related Conditions. Arch Gen Psychiatry 61:361368, 2004
Grilo CM, McGlashan TH, Oldham JM: Course and stability
of personality disorders. Journal of Practical Psychiatry
and Behavioral Health 1:6175, 1998
Grilo CM, McGlashan TH, Morey LC, et al: Internal consistency, intercriterion overlap, and diagnostic efficiency
of criteria sets for DSM-IV personality disorders. Acta
Psychiatr Scand 104:264272, 2001
Grilo CM, Becker DF, Anez LM, et al: Diagnostic efficiency of
DSM-IV criteria for borderline personality disorder: an
evaluation in Hispanic men and women with substance
use disorders. J Consult Clin Psychol 72:126131, 2004a
Grilo CM, Shea MT, Sanislow CA, et al: Two-year stability
and change in schizotypal, borderline, avoidant, and
obsessive-compulsive personality disorders. J Consult
Clin Psychol 72:767775, 2004b
Grilo CM, Sanislow CA, Shea MT: Two-year prospective naturalistic study for remission from major depressive disorder as a function of personality disorder comorbidity.
J Consult Clin Psychol 73:7885, 2005

84

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Gunderson J: Introduction to section IV: personality disorders, in DSM-IV Sourcebook, Vol 2. Edited by Widiger
TA, Frances AJ, Pincus HA, et al. Washington, DC,
American Psychiatric Association, 1996, pp 647664
Gunderson JG, Elliot GR: The interface between borderline
personality disorder and affective disorder. Am J Psychiatry 142:277288, 1985
Gunderson JG, Phillips KA: A current view of the interface
between borderline personality disorder and depression. Am J Psychiatry 148:967975, 1991
Gunderson JG, Morey LC, Stout RL, et al: Major depressive
disorder and borderline personality disorder revisited:
longitudinal interactions. J Clin Psychiatry 65:1049
1056, 2004
Hart AB, Craighead WE, Craighead LW: Predicting recurrence of major depressive disorder in young adults: a
prospective study. J Abnorm Psychol 110:633643, 2001
Herbert JD, Hope DA, Bellack AS: Validity of the distinction between generalized social phobia and avoidant
personality disorder. J Abnorm Psychol 101:332339,
1992
Herkov MJ, Blashfield RK: Clinicians diagnoses of personality disorder: evidence of a hierarchical structure. J Pers
Assess 65:313321, 1995
Hirschfeld RMA, Holzer CE III: Depressive personality disorder: clinical implications. J Clin Psychiatry 55 (suppl):10
17, 1994
Hyler SE: Personality Diagnostic Questionnaire4 (PDQ-4).
New York, New York State Psychiatric Institute, 1994
Hyler SE, Skodol AE, Kellman D, et al: Validity of the Personality Diagnostic Questionnaire: comparison with
two structured interviews. Am J Psychiatry 147:1043
1048, 1990
Hyler SE, Skodol AE, Oldham JM, et al: Validity of the Personality Diagnostic QuestionnaireRevised (PDQ-R): a
replication in an outpatient sample. Compr Psychiatry
33:7377, 1992
Jackson HJ, Burgess PM: Personality disorders in the community: results from the Australian National Survey of
Mental Health and Well-Being, part II. Relationships between personality disorder, Axis I mental disorders,
and physical conditions with disability and health consultations. Soc Psychiatry Psychiatr Epidemiol 37:251
260, 2002
Johnson JG, Cohen P, Skodol AE, et al: Personality disorders
in adolescence and risk of major mental disorders and
suicidality during adulthood. Arch Gen Psychiatry
56:805811, 1999
Johnson JG, Cohen P, Kasen S, et al: Age-related change in
personality disorder trait levels between early adolescence and adulthood: a community-based longitudinal investigation. Acta Psychiatr Scand 102:265275,
2000
Kaplan MA: A womans view of DSM-III. Am Psychol
38:786792, 1983
Karasu TB, Skodol AE: Fourth axis for DSM-III: psychodynamic evaluation. Am J Psychiatry 137:607610, 1980

Kasen S, Cohen P, Skodol AE, et al: The influence of child


and adolescent psychiatric disorders on young adult
personality disorder. Am J Psychiatry 156:15291535,
1999
Kasen S, Cohen P, Skodol AE, et al: Childhood depression
and adult personality disorder: alternative pathways of
continuity. Arch Gen Psychiatry 58:231236, 2001
Kass F, Spitzer RL, Williams JBW: An empirical study of the
issue of sex bias in the diagnostic criteria of DSM-III axis
II personality disorders. Am Psychol 38:799801, 1983
Kass F, Skodol AE, Charles E, et al: Scaled ratings of DSM-III
personality disorders. Am J Psychiatry 142:627630, 1985
Kaye AL, Shea MT: Personality disorders, personality traits,
and defense mechanisms measures, in Handbook of
Psychiatric Measures. Edited by Task Force for the
Handbook of Psychiatric Measures. Washington, DC,
American Psychiatric Association, 2000, pp 713749
Kendell RE, Cooper JE, Gourlay AJ, et al: Diagnostic criteria
of American and British psychiatrists. Arch Gen Psychiatry 25:123130, 1971
Kessler RC, Crum RM, Warner LA, et al: Lifetime co-occurrence of DSM-III-R alcohol abuse and dependence with
other psychiatric disorders in the National Comorbidity
Survey. Arch Gen Psychiatry 54:313321, 1997
Kety SS: Mental illness in the biological and adoptive relatives of schizophrenic adoptees: findings relevant to genetic and environmental factors in etiology. Am J Psychiatry 140:720727, 1983
Kiesler DJ: Contemporary Interpersonal Theory and Research: Personality, Psychopathology, and Psychotherapy. New York, Wiley, 1996
Klein DN: Depressive personality: reliability, validity, and
relation to dysthymia. J Abnorm Psychol 99:412421,
1999
Klein DN: Patients versus informants reports of personality disorders in predicting 7-year outcome in outpatients with depressive disorders. Psychol Assess 15:
216222, 2003
Klein DN, Shih JH: Depressive personality disorder: associations with DSM-III-R mood and personality disorders
and negative and positive affectivity, 30-month stability,
and prediction of course of Axis I depressive disorders.
J Abnorm Psychol 107:319327, 1998
Klonsky ED, Oltmanns TF, Turkheimer E: Informant-reports
of personality disorder: relation to self-reports and future directions. Clin Psychol Sci Pract 9:300311, 2002
Koenigsberg HW, Anwunah I, New AS, et al: Relationship
between depression and borderline personality disorder. Depress Anxiety 10:158167, 1999
Krueger RF, Caspi A, Moffitt TE, et al: The structure and stability of common mental disorders (DSM-III-R): a longitudinal-epidemiological study. J Abnorm Psychol
106:216227, 1998
Krueger RF, Hicks BM, Patrick CJ, et al: Etiologic connections among substance dependence, antisocial behavior, and personality: modeling the externalizing spectrum. J Abnorm Psychol 111:411424, 2002

Manifestations, Clinical Diagnosis, and Comorbidity

Lazarus RS, Folkman S: Stress, Appraisal, and Coping. New


York, Springer, 1984
Lenzenweger MF: Stability and change in personality disorder features: the Longitudinal Study of Personality Disorders. Arch Gen Psychiatry 56:10091015, 1999
Lewinsohn PM, Rohde P, Seeley JR, et al: Axis II psychopathology as a function of Axis I disorder in childhood
and adolescents. J Am Acad Child Adolesc Psychiatry
36:17521759, 1997
Lilienfeld SO, Waldman ID, Israel AC: A critical examination of the use of the term comorbidity in psychopathology research. Clin Psychol Sci Pract 1:7183, 1994
Livesley WJ: Trait and behavioral prototypes of personality
disorder. Am J Psychiatry 143:728732, 1986
Loranger AW: International Personality Disorder Examination (IPDE). Odessa, FL, Psychological Assessment Resources, 1999
Lyons MJ, Tyrer P, Gunderson J, et al: Special feature: heuristic models of comorbidity of Axis I and Axis II disorders. J Personal Disord 11:260269, 1997
McDermut W, Zimmerman M, Chelminski I: The construct
validity of depressive personality disorder. J Abnorm
Psychol 112:4960, 2003
McGlashan TH: Schizotypal personality disorder: Chestnut
Lodge follow-up study, VI. Long-term follow-up perspectives. Arch Gen Psychiatry 43:328334, 1986
McGlashan TH, Grilo CM, Skodol AE, et al: The Collaborative Longitudinal Personality Disorders Study: baseline
patterns of DSM-IV Axis I/II and II/II diagnostic cooccurrence. Acta Psychiatr Scand 102:256264, 2000
McGlashan TH, Grilo CM, Sanislow CA, et al: Two-year prevalence and stability of individual DSM-IV criteria for
schizotypal, borderline, avoidant and obsessive-compulsive personality disorders. Am J Psychiatry (in press)
Millon T, Millon C, Davis R: MCMI-III Manual, 2nd Edition.
Minneapolis, MN, National Computer Systems, 1997
Modestin J, Villiger C: Follow-up study on borderline versus
nonborderline personality disorders. Compr Psychiatry
30:236244, 1989
Moos RH: Coping Responses Inventory. Odessa, FL, Psychological Assessment Resources, 1993
Morey LC, Ochoa ES: An investigation of adherence to diagnostic criteria. J Personal Disord 3:180192, 1989
Morgenstern J, Langenbucher J, Labouvie E, et al: The comorbidity of alcoholism and personality disorders in a clinical population: prevalence rates and relation to alcohol
typology variables. J Abnorm Psychol 106:7484, 1997
Noyes R Jr, Reich J, Christiansen J, et al: Outcome of panic
disorder: relationship to diagnostic subtypes and comorbidity. Arch Gen Psychiatry 47:809818, 1990
Oldham JM, Skodol AE: Charting the future of Axis II. J Personal Disord 14:1729, 2000
Oldham JM, Skodol AE, Kellman HD, et al: Diagnosis of
DSM-III-R personality disorders by two structured interviews: patterns of comorbidity. Am J Psychiatry
149:213220, 1992

85

Oldham JM, Skodol AE, Kellman HD, et al: Comorbidity of


Axis I and Axis II disorders. Am J Psychiatry 152:571
578, 1995
Paris J, Brown R, Nowlis D: Long-term follow-up of borderline patients in a general hospital. Compr Psychiatry
28:530535, 1987
Pearlin LI, Schooler C: The structure of coping. J Health Soc
Behav 19:221, 1978
Perry JC, Cooper SH: An empirical study of defense mechanisms, I: clinical interview and life vignette ratings.
Arch Gen Psychiatry 46:444452, 1989
Pfohl B, Stangl D, Zimmerman M: The implications of DSMIII personality disorders with major depression. J Affect
Disord 7:309318, 1984
Pfohl B, Coryell W, Zimmerman M, et al: DSM-III personality disorders: diagnostic overlap and internal consistency of individual DSM-III criteria. Compr Psychiatry
27:2134, 1986
Pfohl B, Blum N, Zimmerman M: Structured Interview for
DSM-IV Personality. Washington, DC, American Psychiatric Press, 1997
Pope HG Jr, Jonas JM, Hudson JI, et al: The validity of
DSM-III borderline personality disorder: a phenomenologic, family history, treatment response, and long-term
follow-up study. Arch Gen Psychiatry 40:2330, 1983
Reich J: Sex distribution of DSM-III personality disorders in
psychiatric outpatients. Am J Psychiatry 144:485488,
1987
Reich J, Vasile RG: Effect of personality disorders on the
treatment outcome of Axis I conditions: an update.
J Nerv Ment Dis 181:475484, 1993
Reich J, Yates W, Nduaguba M: Prevalence of DSM-III personality disorders in the community. Soc Psychiatry
Psychiatr Epidemiol 24:1216, 1989
Rey JM, Morris-Yates A, Singh M, et al: Continuities between
psychiatric disorders in adolescents and personality
disorders in young adults. Am J Psychiatry 152:895900,
1995
Roberts BW, Caspi A, Moffitt TE: Work experiences and personality development in young adulthood. J Pers Soc
Psychol 84:582593, 2003
Rogers JH, Widiger TA, Krupp A: Aspects of depression associated with borderline personality disorder. Am
J Psychiatry 152:268270, 1995
Sanislow CA, Morey LC, Grilo CM, et al: Confirmatory factor analysis of DSM-IV borderline, schizotypal, avoidant, and obsessive-compulsive personality disorders:
findings from the Collaborative Longitudinal Personality Study. Acta Psychiatr Scand 105:2836, 2002
Seivewright H, Tyrer P, Johnson T: Persistent social dysfunction in anxious and depressed patients with personality
disorder. Acta Psychiatr Scand 109:104109, 2004
Shea MT: Assessment of change in personality disorders, in
Measuring Changes in Mood, Anxiety, and Personality
Disorders: Toward a Core Battery. Edited by Strupp
HH, Horowitz LM, Lambert MJ. Washington, DC,
American Psychological Association, 1997, pp 389400

86

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Shea MT, Widiger TA, Klein MH: Comorbidity of personality disorders and depression: implications for treatment. J Consult Clin Psychol 60:857868, 1992
Shea MT, Pilkonis PA, Beckham E, et al: Personality disorder
and treatment outcome in the NIMH Treatment of Depression Collaborative Research Program. Am J Psychiatry 147:711718, 1990
Shea MT, Stout RL, Gunderson J, et al: Short-term diagnostic
stability of schizotypal, borderline, avoidant, and obsessive-compulsive personality disorders. Am J Psychiatry
159:20362041, 2001
Shedler J, Westen D: Refining personality disorder diagnosis: integrating science and practice. Am J Psychiatry
161:13501365, 2004
Siever LJ, Davis KL: A psychobiological perspective on the
personality disorders. Am J Psychiatry 148:16471658,
1991
Skodol AE: Problems in Differential Diagnosis: From DSMIII to DSM-III-R in Clinical Practice. Washington, DC,
American Psychiatric Press, 1989
Skodol AE, Bender DS: Diagnostic interviews for adults, in
Handbook of Psychiatric Measures. Edited by Task
Force for the Handbook of Psychiatric Measures. Washington, DC, American Psychiatric Association, 2000, pp
4570
Skodol AE, Williams JBW, Spitzer RL, et al: Identifying common errors in the use of DSM-III through diagnostic supervision. Hosp Community Psychiatry 35:251255,
1984
Skodol AE, Rosnick L, Kellman D, et al: Validating structured DSM-III-R personality disorder assessments with
longitudinal data. Am J Psychiatry 145:12971299, 1988
Skodol AE, Oldham JM, Rosnick L, et al: Diagnosis of DSMIII-R personality disorders: a comparison of two structured interviews. Int J Methods Psychiatr Res 1:1326,
1991
Skodol AE, Oldham JM, Hyler SE, et al: Comorbidity of
DSM-III-R eating disorders and personality disorders.
Int J Eating Disorders 14:403416, 1993
Skodol AE, Oldham JM, Gallaher PE, et al: Validity of selfdefeating personality disorder. Am J Psychiatry 151:560
567, 1994
Skodol AE, Oldham JM, Hyler SE, et al: Patterns of anxiety
and personality disorder comorbidity. J Psychiatr Res
29:361374, 1995
Skodol AE, Gallaher PE, Oldham JM: Excessive dependency
and depression: is the relationship specific? J Nerv Ment
Dis 184:165171, 1996
Skodol AE, Oldham JM, Gallaher PE: Axis II comorbidity of
substance use disorders in patients referred for treatment of personality disorders. Am J Psychiatry 156:733
738, 1999
Skodol AE, Gunderson JG, McGlashan TH, et al: Functional
impairment in patients with schizotypal, borderline,
avoidant, or obsessive-compulsive personality disorder. Am J Psychiatry 159:276283, 2002

Skodol AE, Pagano MP, Bender DS, et al: Stability of functional impairment in patients with schizotypal, borderline, avoidant, or obsessive-compulsive personality disorder over two years. Psychol Med (in press)
Snyder S, Sajadi C, Pitts WM Jr, et al: Identifying the depressive border of the borderline personality disorder. Am
J Psychiatry 139:814817, 1982
Soloff PH: Algorithm for pharmacological treatment of personality dimensions: symptom-specific treatments for
cognitive-perceptual, affective and impulsive-behavioral
dysregulation. Bull Menninger Clin 62:195214, 1998
Soloff PH, Ulrich RF: Diagnostic interview for borderline patients: a replication study. Arch Gen Psychiatry 38:686
692, 1981
Somwaru DP, Ben-Porath YS: Development and reliability of
MMPI-2 based personality disorder scales. Paper presented at the 30th annual Workshop and Symposium on
Recent Developments in the Use of the MMPI-2 and
MMPI-A. St. Petersburg, FL, April, 1995
Spitzer RL, Endicott J, Gibbon M: Crossing the border into
borderline personality and borderline schizophrenia:
the development of criteria. Arch Gen Psychiatry 36:17
24, 1979
Sternbach S, Judd A, Sabo A, et al: Cognitive and perceptual
distortions in borderline personality disorder and
schizotypal personality disorder in a vignette sample.
Compr Psychiatry 33:186189, 1992
Stone AA, Schwartz JE, Neale JM, et al: A comparison of
coping assessed by ecological momentary assessment
and retrospective recall. J Pers Soc Psychol 74:1670
1680, 1998
Swartz M, Blazer D, George L, et al: Estimating the prevalence of borderline personality disorder in the community. J Personal Disord 4:257272, 1990
Torgersen S: Genetic and nosological aspects of schizotypal
and borderline personality disorders: a twin study.
Arch Gen Psychiatry 41:546554, 1984
Trull TJ, Widiger TA, Frances A: Covariation of criteria sets
for avoidant, schizoid, and dependent personality disorders. Am J Psychiatry 144:767771, 1987
Turner SM, Beidel DC, Borden JW, et al: Social phobia: Axis
I and II correlates. J Abnorm Psychol 100:102106, 1991
Tyrer P, Gunderson J, Lyons MJ, et al: Special feature: extent
of comorbidity between mental state and personality
disorders. J Personal Disord 11:242259, 1997
Vaillant GE, Bond M, Vaillant CO: An empirically validated
hierarchy of defense mechanisms. Arch Gen Psychiatry
43:786794, 1986
Westen D: Divergences between clinical and research methods for assessing personality disorders: implications for
research and the evolution of Axis II. Am J Psychiatry
154:895903, 1997
Westen D, Moses MJ, Silk KR, et al: Quality of depressive experience in borderline personality disorder and major
depression: when depression is not just depression.
J Personal Disord 6:382393, 1992

Manifestations, Clinical Diagnosis, and Comorbidity

Widiger TA: Personality disorder dimensional models proposed for DSM-V. J Personal Disord 5:386398, 1991
Widiger TA: The DSM-III-R categorical personality disorder
diagnoses: a critique and an alternative. Psychological
Inquiry 4:7590, 1993
Widiger TA: Sex biases in the diagnosis of personality disorders. J Personal Disord 12:95118, 1998
Widiger TA: Personality disorders, in Handbook of Assessment and Treatment Planning for Psychological Disorders. Edited by Antony MM, Barlow DH. New York,
Guilford, 2002, pp 453480
Widiger TA, Coker LA: Assessing personality disorders, in
Clinical Personality Assessment: Practical Approaches,
2nd Edition. Edited by Butcher JN. New York, Oxford
University Press, 2002, pp 407434
Widiger TA, Corbitt EM: Antisocial personality disorder, in
DSM-IV Sourcebook, Vol 2. Edited by Widiger TA,
Frances AJ, Pincus HA, et al. Washington, DC, American Psychiatric Association, 1996, pp 703716
Widiger TA, Frances A: The DSM-III personality disorders:
perspectives from psychology. Arch Gen Psychiatry
42:615623, 1985
Widiger TA, Sanderson CJ: Assessing personality disorders,
in Clinical Personality Assessment: Practical Approaches. Edited by Butcher JN. New York, Oxford University Press, 1995, pp 380394
Widiger TA, Saylor KI: Personality assessment, in Comprehensive Clinical Psychology. Edited by Bellack AS,
Hersen M. New York, Pergamon, 1998, pp 145167
Widiger TA, Hurt SW, Frances A, et al: Diagnostic efficiency
and DSM-III. Arch Gen Psychiatry 41:10051012, 1984
Widiger TA, Trull TJ, Hurt SW, et al: A multidimensional
scaling of the DSM-III personality disorders. Arch Gen
Psychiatry 44:557563, 1987
Widiger TA, Mangine S, Corbitt EM, et al: Personality Disorder InterviewIV: A Semistructured Interview for the
Assessment of Personality Disorders. Odessa, FL, Psychological Assessment Resources, 1995

87

Wiggins JS: Paradigms of Personality Assessment. New


York, Guilford, 2003
Zanarini MC, Gunderson JG, Frankenburg FR: Axis I phenomenology of borderline personality disorder. Compr
Psychiatry 30:149156, 1989
Zanarini MC, Frankenburg FR, Sickel AE, et al: Diagnostic
Interview for DSM-IV Personality Disorders. Belmont,
MA, McLean Hospital, 1996
Zanarini MC, Frankenburg FR, Dubo ED, et al: Axis I comorbidity of borderline personality disorder. Am J Psychiatry 155:17331739, 1998
Zimmerman M: Diagnosing personality disorders: a review
of issues and research methods. Arch Gen Psychiatry
51:225245, 1994
Zimmerman M, Coryell W: DSM-III personality disorder diagnosis in a nonpatient sample: demographic correlates
and comorbidity. Arch Gen Psychiatry 46:682689, 1989
Zimmerman M, Mattia JI: Axis I diagnostic comorbidity and
borderline personality disorder. Compr Psychiatry 40:
245252, 1999a
Zimmerman M, Mattia JI: Differences between clinical and
research practices in diagnosing borderline personality
disorder. Am J Psychiatry 156:15701574, 1999b
Zimmerman M, Pfohl B, Stangl D, et al: Assessment of
DSM-III personality disorders: the importance of interviewing an informant. J Clin Psychiatry 47:261263,
1986
Zimmerman M, Pfohl B, Coryell W, et al: Diagnosing personality disorder in depressed patients: a comparison of
patient and informant interviews. Arch Gen Psychiatry
45:733737, 1988
Zoccolillo M, Pickles A, Quinton D, et al: The outcome of
conduct disorder: implications for defining adult personality disorder and conduct disorder. Psychol Med
22:971986, 1992

This page intentionally left blank

5
Assessment Instruments and
Standardized Evaluation
Wilson McDermut, Ph.D.
Mark Zimmerman, M.D.

Accurate psychodiagnostic assessment of personality


disorders is essential to our understanding of Axis II
pathology. Personality pathology consists of a network of latent constructs for which a taxonomy and
accepted nomenclature already exist. The validity of
the taxonomy, and in turn its theoretical and pragmatic value, are inferred or deduced from the manner
in which we measure personality pathology. Ultimately, our faith in these constructs rests, we would
like to think, on the scientifically established validity
of the assessments we use. Furthermore, the validity
of our measurement instruments is not possible unless
the test is known to be reliable as well.
For the researcher or clinician interested in assessing personality pathology, an array of interviews and
paper-and-pencil tests are available. Most of these instruments measure personality pathology according to
DSM-IV-TR (American Psychiatric Association 2000)
taxonomy, which identifies 10 official personality disorders and 2 appendix (provisional) diagnoses. Although not without its share of controversy and detractors, the DSM-IV-TR personality disorder taxonomy is,

if nothing else, the most widely adopted system for diagnosing personality disorders. However, other conceptualizations of personality pathology have their
own instruments as well. This chapter discusses the interviews and self-administered questionnaires most
widely used in psychiatric research and in the clinical
assessment of personality disorders and pathology.

BACKGROUND
In the bulk of this chapter, we describe the most
commonly used interviews and self-administered
questionnaires for the assessment of personality pathology. Currently, no data conclusively demonstrate
superior reliability or validity for any one structured
interview (Clark and Harrison 2001; Widiger 2002;
Widiger and Coker 2002; Zimmerman 1994). Generally speaking, the assessment instruments described
in the following sections have at least adequate (if not
better) reliability and validity. Some assessment in89

90

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

struments presented may have limited psychometric


data, but the data that do exist are promising. In a few
cases, instruments with limited psychometric data
were included because they represent novel methodologies or are derived from nonDSM-based theories
of personality that are worthy of further systematic
study.
Interested readers should also be aware of several
other thorough and informative reviews of well-known
instruments for the assessment of personality written
by Clark and Harrison (2001), Kaye and Shea (2000),
Rogers (2003), Widiger (2002), Widiger and Coker
(2002), and Zimmerman (1994). Rogers (2003) described the most commonly used interview-based assessments of personality, the importance of incorporating them into routine clinical practice, and reasons for
choosing one instrument over another depending on
the circumstances and the person being interviewed.
The review by Zimmerman (1994) focused only on semistructured interviews, whereas the reviews by Clark
and Harrison (2001), Kaye and Shea (2000), Widiger
(2002), and Widiger and Coker (2002) covered interviews and self-administered questionnaires. These review papers each summarize much of the extant reliability and validity data. They also address pragmatic
concerns such as the who, what, and when of personality assessment and issues critical to interpretability,
such as the effect of co-occurring Axis I disorders on
self-reported personality functioning.

IMPORTANCE OF STANDARDIZED
ASSESSMENTS
Standardized assessments have been developed to
avoid some of the pitfalls of routine, unstructured
clinical interviews (or traditional interviews). Clinical interviews usually begin with questions that focus
on the presenting problem and then usually touch on
several broad areas (e.g., psychiatric history, family
background, psychosocial functioning). Most clinicians adjust their focus throughout the interview and
explore some issues in considerably more detail than
others (Westen 1997). Some clinical settings may have
an intake form, which serves as a rough guideline for
the overall interview. A relatively unstructured interview has the advantage of a high degree of responsiveness to the patients apparent needs and can enhance rapport. Standardized assessments are quite
different. In the case of fully structured interviews, all
questions are provided to the interviewer, who then

reads them verbatim. There is little or no room for departure from the specific set of questions. Semistructured interviews also provide a core set of questions
that are asked in a particular order. Typically, questions tap less threatening areas of functioning first,
then move to material that is less likely to be spontaneously disclosed. In contrast to fully structured interviews, in semistructured interviews the interviewer
has the option of asking follow-up questions to clarify
whether or not a symptom or trait is present. Selfreport questionnaires are equivalent to fully structured interviews that are self-administered (Widiger
2002, p. 463). Almost without exception, standardized
interviews also have highly articulated, systematized
scoring criteria. Standardized assessment procedures
were developed in part because of the poor reliability
of clinical interviews. Poor reliability, typically indexed by lack of agreement between interviewerraters, is a problem for researchers and theorists because it limits validity. In clinical terms, poor reliability means missed diagnoses and misdiagnoses (e.g.,
Rogers 2003; Zimmerman and Mattia 1999a, 1999b).
Standardized assessment procedures for assessing
personality disorders have been de rigueur in research
since the mid-1980s. For reasons discussed further
later in this chapter, it is not recommended that clinicians rely solely on self-report questionnaires to diagnose personality disorders. Regarding the use of
standardized interviews, their adoption into routine
clinical practice has been hindered by several obstacles: perceived detriment to developing rapport due to
the potentially perfunctory nature of conducting interviews, logistical problems, and inadequate training
opportunities. To be sure, a standardized interview
can degenerate into a rapid-fire symptom checklist.
However, when used competently, a standardized interview can provide a reliable and valid assessment.
Research suggests that most diagnostic disagreements
in psychological assessment are not due to the questions but rather to discrepancies in the application of
diagnostic criteria (Widiger and Spitzer 1991). Unstructured interviews and standardized interviews
can coexist. In fact, clinicians could begin an initial interview in an unstructured manner to facilitate rapport and then employ a standardized interview (Rogers 2003).
The main logistical problem in clinical practice is
time. Most clinicians would find it impractical to conduct a standardized interview that can take as long as
2 hours to administer. Having a client complete a selfreport questionnaire first can help narrow the focus of
the interview to those traits and disorders most likely

Assessment Instruments and Standardized Evaluation

to be present and shorten the length of the interview.


Although a personality disorder might not be the
presenting problem, research has established that
patients with personality disorders are less likely to
benefit from treatment for their Axis I symptoms (McDermut and Zimmerman 1997). The last impediment
to the integration of standardized assessment (particularly the interviews) into routine clinical practice is lack
of emphasis in clinical training programs. Doctoral students in clinical psychology often spend a semester
learning projective assessment techniques but get no
exposure to standardized interviews.
Historically, the development of fully structured
and semistructured interviews occurred in research
settings. Westen (1997) pointed out that the process of
diagnosing personality disorders with semistructured
interviews in research settings is very different from
the way clinicians diagnose personality disorders in
routine clinical practice. The main difference is that clinicians do not rely on the direct questions that form
the core of diagnostic instruments. Instead, clinicians
listen to the narratives of patients over time, with
special attention to how the patients describe their interpersonal interactions (Westen 1997). However, although it is true that clinicians arrive at Axis II diagnoses
differently than researchers, it cannot be assumed that
clinical diagnoses are more valid than research diagnoses (Zimmerman and Mattia 1999a), especially
given the research showing the unreliability of unstructured clinical diagnoses (Zimmerman 1994) and
the evidence of validity of research diagnoses (e.g.,
McDermut and Zimmerman 1997).

INTERVIEWS AND CLINICIAN-RATED


INSTRUMENTS
Most of the interviews described in this section are
semistructured. In semistructured interviews, to assess a particular feature the interviewer typically asks
a predetermined question or set of questions. The interviewer can then ask any number of additional questions to clarify what score should be assigned to rate
that feature. The total number of questions in a semistructured interview can be thought of as an approximation of the number of questions one can expect to
ask. However, the actual number of questions can
vary depending on whether the instructions call for
the interviewer to skip certain items or whether the interviewer goes beyond the core questions to ask follow-up questions.

91

Composite International Diagnostic


Interview, Antisocial Section
The Composite International Diagnostic Interview
(CIDI) was developed as part of a collaboration between the World Health Organization and the U.S. Alcohol, Drug Abuse, and Mental Health Administration (Robins et al. 1988). The purpose of this joint
venture, which began in the late 1970s, was to conduct
a cross-national evaluation of the scientific status of alcohol, drug abuse, and mental disorder diagnosis and
classification. In 1982, a task force on instrumentation
began developing diagnostic interviews that would
render diagnoses congruent with widely accepted diagnostic systems such as ICD-10 (World Health Organization 1992) and DSM-III (American Psychiatric Association 1980). One of the final products of the task
force was the CIDI, which was designed primarily for
use with the general population. When the measure
was created, it incorporated questions from the Diagnostic Interview Schedule (DIS; described below), a
semistructured interview already in use. The incorporation of the DIS made the CIDI compatible with
DSM-III. The CIDI covered major clinical syndromes
for the most part, but it also gathered sufficient information to yield a diagnosis of antisocial personality
disorder, which was one of 12 personality disorders in
DSM-III. The CIDI is a fully structured interview in
that questions are read essentially verbatim by examiners, and response options are easily answered by
providing a number or by selecting from among predetermined choices. It is suitable for use by trained
nonclinicians and clinicians alike.

Revised Diagnostic Interview for Borderlines


The Revised Diagnostic Interview for Borderlines
(DIB-R; Zanarini et al. 1989) was designed to assess
Gundersons conceptualization of borderline personality pathology (Gunderson et al. 1981). Gundersons
concept of borderline personality is similar to, but not
identical with, the DSM-IV-TR formulation of borderline personality. The DIB-R is a semistructured interview composed of 105 items that yield ratings on summary statements characterizing borderline pathology.
These summary statements are drawn upon to assess
the following four areas of functioning: impulse action
patterns, affects, cognition, and interpersonal relations. The interview focuses on assessing features
during the past 2 years. The four section scores are
summed to yield a total score ranging from 0 to 10. A
cutoff score of 8 or higher indicates the presence of
borderline personality.

92

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Diagnostic Interview for DSM-IV


Personality Disorders
The Diagnostic Interview for DSM-IV Personality Disorders (DIPD-IV; Zanarini et al. 1996) is a 398-item
structured interview that assesses the 10 DSM-IV-TR
personality disorders as well as passive-aggressive
(negativistic) and depressive personality disorders,
both of which appear in Appendix B of DSM-IV-TR
(Criteria Sets and Axes Provided for Further Study).
Items are grouped by diagnosis. The DIPD-IV determines the presence of traits by focusing on the 2 years
prior to the interview. The DIPD-IV was selected for
use in the Collaborative Longitudinal Personality Disorders Study (Gunderson et al. 2000), the first multisite collaborative study of personality disorders,
which is being conducted in four Northeastern cities.

Diagnostic Interview for Narcissism


The Diagnostic Interview for Narcissism (Gunderson
et al. 1990) assesses narcissism as it is conceptualized
by Gunderson and colleagues. Narcissism in this view
is more heterogeneous than in DSM-IV-TRs conceptualization. The Diagnostic Interview for Narcissism
generates ratings of grandiosity, interpersonal relations, reactiveness, affects and mood states, and social
and moral judgments. The ratings in these areas are
derived from an interview composed of 33 statements. For clinicians and researchers interested in the
assessment of narcissism and narcissistic personality
disorder, Hilsenroth et al. (1996) provide an excellent
review of extant instruments.

Hare Psychopathy ChecklistRevised


The Hare Psychopathy ChecklistRevised (PCL-R;
Hare 1991) can be used to assess psychopathy both
categorically and dimensionally. It was designed for
use primarily in forensic settings. The construct of
psychopathy is somewhat broader than in DSM-IVTR, in which the definition of psychopathy consists
predominantly of a history of antisocial behaviors.
Psychopathy also encompasses glibness and charm,
grandiosity, lack of empathy, and shallow affect. The
PCL-R is a 20-item checklist; thus, it is not strictly an
interview. However, information for rating of items
can be gleaned from a semistructured interview and/
or ancillary sources (e.g., institutional records). Clinical judgment and inference are required for scoring
most of the items. Items are scored from 0 to 2, in
which a 2 indicates that the item is true of the examinee. The total score ranges from 0 to 40. The scale has

two subscales (or factors); Factor 1 represents psychopathic personality characteristics, and Factor 2
represents socially deviant behaviors. Scoring the
PCL-R involves generating scores for each of the factors and a combined total score. A cutoff score of 30 or
greater can be used to signify the presence of psychopathy.

International Personality Disorders


Examination
The International Personality Disorders Examination
(IPDE; Loranger 1999) is a 537-question semistructured interview that evaluates personality disorders
according to both DSM-IV-TR and ICD-10 criteria.
Personality disorders included in the ICD-10 are as
follows: paranoid, schizoid, dissocial, emotionally unstableimpulsive, emotionally unstableborderline,
histrionic, anankastic, anxious, and dependent. The
interviewer inquires about age of onset of pathologic
traits, and at least one trait must have been present before age 25 years. The IPDE is designed for use by professionals with substantial psychodiagnostic experience. The IPDE questions are organized by topic:
work, self, interpersonal relationships, affects, reality
testing, and impulse control. The IPDE has been translated into several different languages for use in a multisite international study of personality disorders
(Loranger et al. 1994). The IPDE also has a 77 true/
false question screener that is completed by the subject
prior to the interview.

National Institute of Mental Health


Diagnostic Interview Schedule,
Antisocial Section
The Diagnostic Interview Schedule (DIS; Robins et al.
1981) was developed at the request of the National
Institute of Mental Health for use in the Epidemiologic Catchment Area projects (Regier et al. 1984). Its
structure and features followed the general design of
the Renard Diagnostic Interview (Helzer et al. 1981),
which was used to make diagnoses consistent with
the Washington University criteria (Feighner et al.
1972). Features of the Renard Diagnostic Interview
that were incorporated into the DIS were that all
questions and probes were fully specified and that
diagnoses were made according to a computer algorithm to minimize clinical judgment. These features,
it was hoped, would allow lay interviewers with 12
weeks of training to make diagnoses as accurately as
psychiatrists. The use of lay interviewers was consid-

Assessment Instruments and Standardized Evaluation

ered important to avoiding the high cost and impracticality of employing psychiatrists as interviewers in
large epidemiologic studies. The DIS gathered information primarily on mood anxiety, substance, and
psychotic disorders and generated diagnoses according to multiple diagnostic systems, including the
DSM-III. The only personality disorder measured
was antisocial personality disorder.

Personality Assessment Form


The Personality Assessment Form (Pilkonis et al. 1991)
is not an interview per se, but could best be characterized as a clinician-report instrument (Widiger 2002).
It does not provide of list of questions to ask a subject,
nor does it yield categorical assignments of personality disorder diagnoses. Rather, it provides a brief description of the important features of each personality
disorder and a six-point scale against which to make a
diagnosis. Thus, the form requires substantial clinical
judgment. Based on previous research, a cutoff score
of four or higher has been established as the threshold
for identifying cases of personality disorder (Shea et
al. 1990).

Personality Assessment Schedule


The Personality Assessment Schedule, developed by
Tyrer (1988) in Britain, is a comprehensive interview
that assesses 24 traits (e.g., conscientiousness, aggression, and impulsivity) and generates dimensional ratings of five personality styles: normal, passive-dependent, sociopathic, anankastic (analogous to obsessivecompulsive), and schizoid. Regrettably, this instrument has received little attention in the United States,
despite being a comprehensive interview that yields
dimensional ratings of personality style.

Personality Disorder InterviewIV


The Personality Disorder InterviewIV (PDI-IV; Widiger et al. 1995) is the most current version of what
was previously known as The Personality Interview
Questions (versions I, II, and III). The PDI-IV has
questions assessing the 94 DSM-IV-TR criteria for the
10 official DSM-IV-TR personality disorders and two
appendix diagnoses. A trait is rated as present if it
has been characteristic for much of the subjects adult
life and present since age 18. The PDI-IV has two versions, one with items grouped by diagnosis and the
other with items grouped by topic. A translated version of the PDI-IV was recently used in China (Yang
et al. 2000).

93

Structured Clinical Interview for DSM-IV


Axis II Personality Disorders
The Structured Clinical Interview for DSM-IV Axis II
Personality Disorders (SCID-II; First et al. 1997) is a
119-item semistructured interview with items keyed
to the DSM-IV-TR personality disorder criteria. The
SCID-II evaluates traits for the past 5 years. The presence of traits and disorders is operationalized based
on DSM-IV-TRs guideline that subjects describe how
they have generally felt, exclusive of Axis I symptoms.
The essential features of each disorder should have
been present cross-contextually since early adulthood.
Authors of the SCID-II recommend administration
only by experienced clinicians. Items are grouped by
diagnosis. The SCID-II interview is preceded by administration of the SCID-II Personality Questionnaire.
Interviewers then follow up on items endorsed as
present by the subject on the screener. The screener
has 119 items.

Structured Interview for DSM-IV Personality


Disorders
The Structured Interview for DSM-IV Personality Disorders (SIDP-IV; Pfohl et al. 1997) is a 101-item semistructured interview that assesses the 10 DSM-IV-TR
official personality disorders plus the proposed depressive, self-defeating, and negativistic personality disorders. Each item is rated from 0 to 3 (0=not present,
1=subthreshold, 2=present, 3=strongly present). The
items are keyed to the DSM-IV-TR personality disorder
criteria. The SIDP comes in two versions: one in which
items are grouped topically and another in which items
are grouped by diagnosis. The interviewee is asked to
focus on his or her usual self; and if there has been a
dramatic recent change in the individuals personality,
then the functioning that predominated for the greatest
amount of time in the past 5 years is considered typical.
For a diagnosis to be considered present, the traits endorsed must have been present for the majority of time
in the past 5 years. The SIDP is designed for use by individuals with a minimum of a bachelors degree in the
social sciences, 6 months experience interviewing psychiatric patients, and about 1 month of specific training
in using the SIDP.

Structured Interview for the Five-Factor


Model of Personality
The 120-item Structured Interview for the Five-Factor
Model of Personality (Trull and Widiger 1997) is
unique in that it is the only semistructured interview

94

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

that assesses general personality. It is modeled after


the NEO Personality InventoryRevised (NEO-PI-R;
Costa and McCrae 1992) in that it assesses the five domains of the five-factor model (FFM) of personality,
which include neuroticism, extraversion, openness to
experience, agreeableness, and conscientiousness. It
also assesses the six facets of each of the five major domains. The Structured Interview for the Five-Factor
Model of Personality has a slightly stronger emphasis
on maladaptive components of general personality
than does the NEO-PI-R.

SELF-ADMINISTERED QUESTIONNAIRES
Coolidge Axis II Inventory
The Coolidge Axis II Inventory (CATI; Coolidge and
Merwin 1992) has 200 items, rated on a four-point
true/false scale ranging from 1 (strongly false) to 4
(strongly true). The personality disorder items were
selected or developed specifically to assess the DSM
Axis II symptoms. In addition to assessing 13 DSMIII-R personality disorders (11 official personality disorders plus sadistic and self-defeating personality
disorders as described in the DSM-III-R appendix)
(American Psychiatric Association 1987), the CATI
also has scales to assess depression, anxiety, and brain
dysfunction.

Dimensional Assessment of Personality


PathologyBasic Questionnaire
The Dimensional Assessment of Personality PathologyBasic Questionnaire (DAPP-BQ; Livesley and
Jackson, in press) is a 290-item assessment instrument
that assesses 18 dimensions of personality pathology.
Respondents rate each item on a five-point Lykert-type
scale in which a score of 1 equals very unlike me and
a score of 5 equals very like me. Items included were
those that highlighted traits and behavioral acts characteristic of DSM-III personality disorders, but items were
not explicit paraphrasings of DSM-III personality disorder criteria. Examples of some of the dimensions assessed include some that correspond to DSM personality disorder criteria (e.g., self-harming behaviors, social
avoidance); some that correspond to prototypical features of particular personality disorders (e.g., narcissism); some that cover interpersonal difficulties (e.g.,
intimacy problems); and some that span both the traits
traditionally studied in academic psychology and disordered personality, which has been the traditional emphasis of psychiatric research on personality (e.g.,

anxiousness, compulsivity). Although the DAPP-BQ


covers the components of DSM personality disorders, it
does not render scale scores that correspond with them.

Inventory of Interpersonal Problems


The Inventory of Interpersonal Problems is a 64-item
self-administered questionnaire (Horowitz et al. 2000).
The items assess a wide range of interpersonal problems. Respondents rate items in terms of how distressing the problem has been, ranging from 0 (not at all) to
4 (extremely). The interpersonal theory on which the
scale is based is an adaptation of the interpersonal circumplex (IPC) model of interpersonal dispositions.
According to this model, interpersonal behavior can be
located in two-dimensional circular space, with dominance versus submission on one axis and hostility versus friendliness on the other axis. The scale yields information about a persons interpersonal behavior
with respect to the following areas: being domineering,
vindictive, cold, avoidant, unassertive, exploitable, hypernurturing, and intrusive. A 32-item short form is
also available.

Millon Clinical Multiaxial InventoryIII


The Millon Clinical Multiaxial InventoryIII (MCMIIII; Millon et al. 1997) is a 175-item true/false questionnaire that assesses Axis I and II pathology. Now in its
third generation, this inventory has been one of the
most widely used paper-and-pencil tests in research
and the most widely used paper-and-pencil test employed clinically to generate actual diagnoses. Its purpose is to operationalize the assessment of Millons
theory of psychopathology. Millons proposed psychopathologic constructs are congruent to a great degree, but not entirely, with the disorders in DSM-IVTR. The MCMI-III assesses many Axis I disorders and
the following personality disorders in addition to
those assessed by the DSM-IV-TR: aggressive (sadistic), self-defeating (masochistic), depressive, and negativistic (passive-aggressive).

Minnesota Multiphasic Personality


InventoryPersonality Disorder Scales
The Minnesota Multiphasic Personality Inventory
(MMPI) Personality Disorder Scales (Morey et al.
1985) were developed using a two-step, rationalempirical process. In the first step, clinical psychologist judges searched the 566 MMPI items and selected
those expected to be representative of one or more of
the 11 DSM-III personality disorders. In the second

Assessment Instruments and Standardized Evaluation

step, the selected items were analyzed iteratively to


determine that they discriminated between high and
low scorers on the scales to which they belonged. The
result was 154 true/false items. Some items were common to two or more scales, so the authors developed
two sets of scales, one in which items overlapped and
a second set with no overlapping items. These personality disorder scales have been updated (MMPI-II,
Colligan et al. 1994), but have not yet been coordinated with DSM-IV-TR.

Minnesota Multiphasic Personality


InventoryPersonality and
Psychopathology Five Scales
Using MMPI items, Harkness et al. (1995) developed
five scales (the PSY-5) to facilitate the description of
general personality and to complement the diagnosis of
personality disorders. The five constructs these scales
measure are aggressiveness, psychoticism, constraint,
negative emotionality/neuroticism, and positive emotionality/extraversion.

Narcissistic Personality Inventory


The Narcissistic Personality Inventory (Raskin and
Terry 1988) is a 40-item self-administered questionnaire that measures trait narcissism. Although many
items were originally constructed to correspond to
features of DSM-III narcissistic personality disorder,
the instrument is not intended to yield categorical
diagnoses. Each item on the inventory consists of a
pair of statements (one that reflects narcissism, the
other nonnarcissistic). The respondent is instructed to
choose the item that is most true.

NEO Personality InventoryRevised


The NEO-PI-R (Costa and McCrae 1992) is designed to
assess general personality traits according to the FFM.
The development of this instrument grew out of academic psychologys traditional interest in normal personality traits and dimensions. The five factors are
thought to be fundamental and nearly ubiquitous dimensions of personality, representing higher-order
traits composed of multiple lower-order traits. In addition to the putative universality of the five factors,
the theory holds that individual differences in the expression of personality can be explained in terms of
any given individuals location along each of the five
basic dimensions. The NEO-PI-R is a 240-item selfadministered questionnaire intended to assess the five

95

factors (see earlier discussion, Structured Interview


for the Five-Factor Model of Personality). Within
each higher-order factor, the NEO-PI-R also assesses
six lower-order traits or facets. For example, the six
facets of the higher-order trait of neuroticism that are
measured by the NEO-PI-R are depression, anxiety,
angry hostility, self-consciousness, vulnerability, and
impulsiveness. There is also a 60-item alternative to
the NEO-PI-R, the NEOFive Factor Inventory (Costa
and McCrae 1992), which is designed to assess just the
five factors of the FFM, not their constituent facets.

Personality Diagnostic Questionnaire4


The Personality Diagnostic Questionnaire4 (Hyler
1994) produces diagnoses consistent with DSM-IV-TR
criteria for the 10 official and 2 appendix Axis II diagnoses. It consists of 85 items. Previous versions have
shown adequate test-retest reliability. Items were selected or developed specifically to assess DSM Axis II
symptoms. The questionnaires predecessors had high
sensitivity but lower specificity. It generated many
false positive diagnoses, but very few false negatives.
Yang et al. (2000) used a Chinese version of the Personality Diagnostic Questionnaire4 in a large sample of
psychiatric patients in the Peoples Republic of China.

Personality Assessment Inventory


The Personality Assessment Inventory is a self-report
questionnaire consisting of 344 items (Morey 1991). It
produces 22 scales that provide continuous ratings of
major clinical syndromes, personality features, and
factors that may compromise treatment. Among the
22 scales are also 4 validity scales. Each item is rated
by respondents on a 4-point Lykert-type scale. The
Personality Assessment Inventory has four personality dimensions including borderline features, antisocial features, and the interpersonal dimensions of
dominance and warmth.

Psychopathic Personality Inventory


The Psychopathic Personality Inventory (Lilienfield
and Andrews 1996) is a self-administered questionnaire that was developed to assess traits of psychopathy (i.e., the assessment of antisocial acts is deemphasized). The measure contains eight subscales developed
using factor analysis: Machiavellian egocentricity, social potency, coldheartedness, carefree nonplanfulness,
fearlessness, blame externalization, impulsive nonconformity, and stress immunity. The Psychopathic Per-

96

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

sonality Inventory also contains validity scales to detect


malingering and careless responding.

Schedule for Nonadaptive and


Adaptive Personality
The Schedule for Nonadaptive and Adaptive Personality (SNAP) contains 375 true/false questions (Clark
1993). Many items were selected or developed specifically to assess DSM Axis II symptoms. In addition to
12 diagnostic scales that are congruent with DSM Axis
II constructs, the SNAP has validity scales and 15 trait
and temperament scales. The items reflect a broad array of personality disorder descriptors, corresponding
in some cases to the DSM. Other items were based on
non-DSM formulations of personality disorder, and
others congruent with symptoms of particular Axis I
disorders with traitlike manifestations (e.g., chronic
disturbances associated with dysthymia or generalized anxiety disorder). Scoring rules for making DSMIV-TR personality disorder diagnoses have also been
established.

Shedler-Westen Assessment Procedure


The Shedler-Westen Assessment Procedure (SWAP200; Westen and Shedler 1998) is composed of 200
items representing the 94 DSM-IV-TR personality disorder diagnostic criteria, personality disorder symptomatology, defense mechanisms, and adaptive personality traits. The respondent is a clinician who rates
the patient using a Q-sort procedure, in which he or
she rates items according to a Lykert-style format in
which ratings conform to a predetermined distribution. For example, Westen and Shedler (1998) used an
eight-point scale (from 0 [not at all descriptive, irrelevant, or inapplicable] to 7 [highly descriptive]), and
clinician raters were required to assign a rating of 7 on
eight of the 200 items. Ten items were required to receive a score of 6, and 100 of the 200 items were required to receive a rating of 0.

Schizotypal Personality Questionnaire


The Schizotypal Personality Questionnaire (Raine 1991)
is a 74-item self-administered questionnaire with items
assessing each of the DSM-III-R schizotypal personality
disorder criteria. The questionnaire was developed to
assess schizotypal personality patterns and to screen
for schizotypal personality disorder in the community.
Raine and Benishay (1995) developed a brief 22-item
version to be used as a screening instrument.

Structural Analysis of Social Behavior


Intrex Questionnaire
The Structural Analysis of Social Behavior Intrex Questionnaire is a series of questionnaires that operationalize concepts outlined in Benjamins Structural Analysis
of Social Behavior (SASB) model (Benjamin 1996). The
SASB model delineates three important aspects of
inter- and intrapersonal behavior: focus, affiliation, and
interdependence. It has roots in Learys (1957) IPC
model and in the work of Sullivan (1953). Similar to the
IPC, the SASB model depicts a horizontal friendlinessversus-hostility (affiliation) axis. Unlike the IPC, in the
SASB model the affiliation axis is crossed with a vertical enmeshment-versus-differentiation axis, called
interdependence. Enmeshment refers to control and submission. Control and submission are not depicted as diametrically opposed, but rather they are conceived of
as complementary, differing only in terms of their focus of action, where control is directed toward another
and submission is in response to another. Differentiation
refers to processes called emancipation, separation, and
assertiveness. In the SASB model, there are three IPCs
(with affiliation and interdependence axes), one for
each of three foci of action: transitive, intransitive, and
introjective. Transitive, intransitive, and introjective refer
to whether or not social action is toward others (transitive), in reaction to others (intransitive), or toward oneself (introjective).
The SASB consists of a series of self-administered
questionnaires, selected by the patient and clinician in
collaboration. There is a standard series comprising
versions directed toward self, significant other, mother,
father, mother in relationship with father, and father in
relationship with mother. Patients are asked to rate
themselves or others at their best and at their worst.
Items are rated from 0 to 100 in 10-point increments,
ranging from 0 (never, or not at all applicable) to 100
(always, perfectly applicable). Up to 36 scores are
plotted for each of the three IPCs, representing different
foci of action. Numerous other scores, which are often
generated by complex mathematical algorithms, are
available.

Temperament and Character Inventory


The Temperament and Character Inventory (TCI;
Cloninger et al. 1994) is a 240-item self-administered
questionnaire designed to measure personality from
the perspective of Cloningers seven-factor model
(Cloninger et al. 1993). The TCI measures four dimensions of temperament and three dimensions of character. Cloninger et al. (1993) postulated that temperament

Assessment Instruments and Standardized Evaluation

is highly heritable, remains stable throughout life, and


has specific neurobiological and neuroanatomical substrates. The character dimensions, on the other hand,
are more malleable. Character is postulated to be modifiable through learning and sociocultural influences
and is capable of evolving throughout the lifespan.
Cloningers model contains four dimensions of temperament (novelty seeking, harm avoidance, reward
dependence, and persistence) and three dimensions of
character (self-directedness, cooperativeness, and selftranscendence) (see Chapter 9, Genetics).

Wisconsin Personality Inventory


The Wisconsin Personality Inventory (Klein et al.
1993) is a 214-item self-administered questionnaire
that yields dimensional as well as categorical scores
for DSM-III-R personality disorders. It contains items
coordinated with each personality disorder diagnostic
criterion. Each item is rated on a 10-point scale from
1 (never, not at all) to 10 (always, extremely). Respondents are asked to focus on what is true of them during the past 5 years or more. Many of the items were
written from the interpersonal, object relational standpoint of Benjamins (1996) SASB model.

ADVANTAGES AND DISADVANTAGES OF


PERSONALITY ASSESSMENT INSTRUMENTS
Semistructured Interviews
None of the assessment instruments described earlier
has been shown to have unequivocally superior reliability and validity, and no instrument is without distinct disadvantages. Advantages of various assessment instruments are discussed later. Among the
available semistructured interviews, one obvious consideration is whether the interview assesses all DSMIV-TR personality disorders. There are five semistructured interviews developed specifically to correspond
to DSM-IV-TR personality disorders: DIPD-IV, IPDE,
PDI-IV, SCID-II, SIDP-IV. For clinical purposes, the
IPDE and PDI-IV in particular have detailed administration and scoring manuals that can be valuable assets for clinicians. The IPDE, SCID-II, and SIDP-IV
have been used in the most empirical studies. The
IPDE and SIDP-IV also provide information about
which questions are required to make diagnoses according to ICD-10. The IPDE is the longest to administer, with 537 questions requiring up to 2 hours. The
IPDE and SCID-II have screening questionnaires that
can help narrow the focus to traits and disorders most

97

likely to be present. However, a clinician might want


to use one of the self-report questionnaires described
earlier as a screening device, because there are many
questionnaires whose psychometric properties are
supported by much larger bodies of research. If a clinician wants to focus the interview on traits and/or
disorders identified by a screening questionnaire, a
standardized interview with questions arranged by
disorder will be easier to use because finding the relevant questions will be easier. All standardized interviews except the IPDE have versions organized in a
disorder-by-disorder format. The IPDE, PDI-IV, and
SIDP-IV also have versions organized by thematic
content (e.g., interpersonal relations, work, interests,
and hobbies). The thematic organization of items is
thought to mitigate against potential halo effects.
In addition to comprehensive semistructured interviews for the assessment of all DSM-IV-TR personality disorders, other interviews specifically target the
assessment of borderline pathology, narcissism, and
psychopathy. The DIB-R and Diagnostic Interview for
Narcissism assess borderline pathology and narcissism, respectively, from Gundersons conceptualization (Gunderson et al. 1981, 1990). These interviews
have the advantage of furnishing rich descriptions of
an individuals functioning in these areas, but they can
be almost as time consuming as the comprehensive
semistructured interviews covering all the DSM personality disorders. The PCL-R is a measure of Hares
concept of psychopathy, which in addition to gathering data about an individuals history of antisocial behavior includes coverage of features such as glibness,
superficiality, and charm associated with psychopathy. The Structured Interview for the Five-Factor
Model of Personality can be used to flesh out the description of a clients general personality functioning
(e.g., extraversion vs. introversion; antagonism vs.
agreeableness) from the standpoint of the FFM.

Self-Report Questionnaires
There are many advantages to using self-report questionnaires as well, although all are problematic for
varying reasons. If one wants a comprehensive measure of personality disorders from the DSM perspective, there are several options: MMPI personality
disorder scales, MCMI-III, Personality Diagnostic
Questionnaire4, and CATI. The MMPI personality
disorder scales and the CATI have the disadvantage
of not being coordinated with DSM-IV. The MMPI
scales are also embedded within the 567 items of the
MMPI-I, which can be time-consuming to complete.
The MCMI-III has been the most heavily researched

98

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

(see Craigs 1999 review of MCMI research), although


research suggests that the MCMI-III is prone to gender bias, and the technical manual does not explain in
sufficient detail the mathematics underlying the determination of base rates. Research indicates that selfreport questionnaires assessing personality disorders
tend to detect many false positives compared with interviews. However, they are highly sensitive, making
self-report measures useful as screening devices. Selfreport measures have also been shown to demonstrate high convergent validity, probably as a result of
being so structured (Widiger 2002).
There are also self-report measures of personality
functioning that grew out of nonDSM-based theoretical and or research traditions. Although the MCMI-III
provides scores on DSM-IV-TR personality disorders,
Millons theory of psychopathologyrooted in evolutionary theoryprovides ratings of personality pathology (depressive, negativistic, masochistic, sadistic) that
are not officially part of DSM-IV-TR. The Wisconsin
Personality InventoryIV, Inventory of Interpersonal
Problems, and SASB are heavily influenced by interpersonal theory. The Wisconsin Personality Inventory
IV and SASB also have strong ties to object relations
theory. The SNAP and DAPP-BQ provide ratings of
personality functioning that complement but are by no
means identical to the DSM formulation of personality
disorders. The Wisconsin Personality InventoryIV,
Inventory of Interpersonal Problems, SASB, SNAP, and
DAPP-BQ have not been as heavily researched as the
scales from the MMPI, MCMI-III, and Personality Diagnostic Questionnaire4. The TCI measures personality from the perspective of a comprehensive psychobiological approach to personality functioning. For
comprehensive evaluations of personality functioning
from the standpoint of general personality, there are the
NEO-PI-R and the NEO Five-Factor Model. The NEOPI-R has been found to provide valuable information
above and beyond pathology-laden assessment techniques based on DSM models (Garb 2003). The PSY-5 is
a recently developed set of five subscales of the MMPIII that provide scores on measures of both general personality functioning and personality pathology. There
are also innovative approaches to rating personality
functioning such as the SWAP-200, which uses a clinician-rated Q-sort methodology. The SWAP-200 also
generates ratings of defense mechanisms, although ratings must adhere to a specified distribution and thus
may result in incomplete coverage of an individuals
personality functioning. For assessments that target
one or two dimensions of personality pathology, there

are the Personality Assessment Inventory (borderline,


antisocial), Psychopathic Personality Inventory (psychopathy), Narcissistic Personality Inventory (narcissism), and Schizotypal Personality Questionnaire
(schizotypy). The Psychopathic Personality Inventory,
Narcissistic Personality Inventory, and Schizotypal
Personality Questionnaire are relatively short and thus
are useful for targeted investigations of specific dimensions of functioning. The borderline and antisocial
scales in the Personality Assessment Inventory are embedded in a larger 344-item questionnaire, which may
be impractical for certain uses. It also has yet to be coordinated with DSM-IV-TR. However, the Personality
Assessment Inventory has multiple subscales capturing Axis I pathology, validity scales, and measures of
personality dimensions related to dominance and
warmth.

IMPORTANT ISSUES IN THE ASSESSMENT OF


PERSONALITY PATHOLOGY
Effect of Axis I Symptoms on
Reported Personality
It is now well established that Axis I symptoms, such
as acute depressive, anxious, or psychotic states, can
bias the self-reported personality characteristics of patients (Hirshfeld et al. 1983; Piersma 1989; Zimmerman 1994). Depressed patients, for example, will depict themselves in a more negative light (introverted,
dependent, inadequate) than they would have in a
nondepressed state (Widiger 1993). However, individuals with eating disorders (Ames-Frankel et al. 1992)
and obsessive-compulsive disorder (Ricciardi et al.
1992) have also been shown to report lower levels of
personality pathology following treatment, relative to
reported levels of personality pathology at treatment
initiation. It may be tempting to think that semistructured interviews can circumvent the state-biasing effect of acute Axis I symptoms on reported personality
pathology. The comprehensive semistructured interviews make a point of trying to distinguish the patients usual personality from personality functioning
at the initiation of treatment for Axis I disorders. On
balance, however, the available evidence suggests that
both personality disorder interviews and self-report
questionnaires are prone to overreporting bias due to
psychiatric state (Widiger and Coker 2002; Zimmerman 1994).

Assessment Instruments and Standardized Evaluation

Use of Informants
Most information about personality pathology comes
directly from patients and thus is vulnerable to distortions or omissions that could undermine the validity of
the material provided. As previously discussed, reported traits may be affected by comorbid Axis I pathology. Alternatively, patients may deny the presence
of socially undesirable behavior, lack insight, or simply
be unaware of the effect their behavior has on others.
Thus, there has been increased research on the degree
to which informants (e.g., spouses, relatives, friends)
can elucidate the presence of personality pathology in
patients. From the standpoint of the researcher, obtaining consistent data from patients and informants
serves as a form of convergent validity (Widiger 1993).
Overviews of the state of the research on patientinformant agreement have generally noted poor to adequate agreement (Widiger and Coker 2002).
Despite the low to modest agreement between patients and informants, most researchers have stressed
the value of including informants in Axis II assessments
whenever feasible (Clark and Harrison 2001; Widiger
2002). Even when patient and informant reports do not
agree at all on the presence of a personality disorder, informants may be identifying patients with personality
disorders who did not self-identify in their own interviews. For example, Riso et al. (1994) pointed out that of
all the personality disorder diagnoses made based on
informant interview, only 18% of those disorders were
also made based on patient interviews. From the standpoint of the clinician, the use of information from a collateral source (such as a significant other) may illuminate personality traits that the patient denies or is
unaware of and may help clinicians mitigate the biasing
effect of Axis I symptomatology on the patients selfreport.

INTERVIEW VERSUS QUESTIONNAIRE


There are fewer logistical problems associated with
conducting studies that compare interviews and selfreport measures, and consequently these types of
comparisons are conducted far more frequently (Zimmerman 1994). In both patient and nonpatient samples, researchers have found low levels of agreement
between interviews and questionnaires (i.e., mean
kappas between 0.25 and 0.36) (Zimmerman 1994).
The reason for the low agreement is primarily the fact
that questionnaires tend to overdiagnose. In some

99

cases, questionnaires overdiagnose the presence of a


personality disorder at a rate almost 10 times higher
than that of interviews (Hunt and Andrews 1992). The
high sensitivity of questionnaires suggests that they
may be useful as screening measures in clinical settings but are inappropriate for use in making diagnoses.
Questionnaires also have value for researchers. Assuming structured interviews demonstrate reliability,
there still remains the possibility of intersite differences in terms of the unstructured follow-up questions that might be used or the interpretation of diagnostic criteria. Therefore, Zimmerman et al. (1993)
proposed that self-report questionnaires be used as a
paper standard. In other words, if two research centers
obtain different prevalence rates for personality disorders, the validity of their respective findings could be
judged against the degree of concordance between interview results and questionnaire results at the two
centers. Assuming no real population differences, the
questionnaire thus becomes the definitive standard.

REFERENCES
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition. Washington,
DC, American Psychiatric Association, 1980
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition, Revised. Washington, DC, American Psychiatric Association, 1987
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition. Washington,
DC, American Psychiatric Association, 1994
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association, 2000
Ames-Frankel J, Devlin MJ, Walsh BT, et al: Personality disorder diagnoses in patients with bulimia nervosa: clinical correlates and changes with treatment. J Clin Psychiatry 53:9096, 1992
Benjamin LS: Interpersonal Diagnosis and Treatment of Personality Disorders, 2nd Edition. New York, Guilford,
1996
Clark LA: Manual for the Schedule for Nonadaptive and
Adaptive Personality. Minneapolis, MN, University of
Minnesota Press, 1993
Clark LA, Harrison JA: Assessment instruments, in Handbook of Personality Disorders. Edited by Livesley WJ.
New York, Guilford, 2001, pp 277306
Cloninger CR, Svrakic DM, Przybeck TR: A psychosocial
model of temperament and character. Arch Gen Psychiatry 50:975990, 1993

100

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Cloninger CR, Przybeck TR, Svrakic DM, et al: The Temperament and Character Index: A Guide to Its Development and Use. St Louis, MO, Washington University,
Center for Psychobiology of Personality, 1994
Colligan RC, Morey LC, Offord KP: MMPI/MMPI-2 personality disorder scales: contemporary norms for adults and
adolescents. J Clin Psychol 50:168200, 1994
Coolidge FL, Merwin MM: Reliability and validity of the
Coolidge Axis II Inventory: a new inventory for the assessment of personality disorders. J Pers Assess 59:223
238, 1992
Costa PT, McCrae RR: Revised NEO Personality Inventory
(NEO PI-R) and NEO Five Factor Inventory Professional Manual. Odessa, FL, Psychological Assessment
Resources, 1992
Craig RJ: Overview and current status of the Millon Clinical
Multi-axial Inventory. J Pers Assess 72:390406, 1999
Feighner JP, Robins E, Guze SB, et al: Diagnostic criteria for
use in psychiatric research. Arch Gen Psychiatry 26:57
63, 1972
First MB, Gibbon RL, Spitzer RL, et al: Structured Clinical Interview for DSM-IV Axis II Personality Disorders, (SCIDII). Washington, DC, American Psychiatric Press, 1997
Garb H: Incremental validity and the assessment of psychopathology in adults. Psychol Assess 15:508520, 2003
Gunderson JG, Kolb JE, Austin V: The Diagnostic Interview
for Borderline Patients (DIB). Am J Psychiatry 138:896
903, 1981
Gunderson JG, Ronningstam E, Bodkin A: The Diagnostic
Interview for Narcissistic Patients. Arch Gen Psychiatry
47:676680, 1990
Gunderson JG, Shea MT, Skodol AE, et al: The collaborative
longitudinal personality disorders study: development,
aims, design, and sample characteristics. J Personal Disord 14:300315, 2000
Hare RD: The Hare Psychopathy ChecklistRevised Manual. North Tonawanda, NY, Multi-Health Systems, 1991
Harkness AR, McNulty JL, Ben-Porath YS: The Personality
Psychopathology Five (PSY-5): constructs and MMPI-2
scales. Psychol Assess 7:104114, 1995
Helzer JE, Robins LN, Croughan JL, et al: Renard Diagnostic
Interview: its reliability and procedural validity with
physicians and lay interviewers. Arch Gen Psychiatry
38:393398, 1981
Hilsenroth MJ, Handler L, Blais MA: Assessment of Narcissistic Personality Disorder: a multi-method review. Clin
Psychol Rev 16:655683, 1996
Hirshfeld RMA, Klerman GL, Clayton PJ, et al: Assessing
personality: effects of the depressive state on trait measurement. Am J Psychiatry 140:695699, 1983
Horowitz LM, Alden LE, Wiggins JS, et al: Inventory of Interpersonal Problems: IIP-64. San Antonio, TX, Harcourt Assessment, 2000
Hunt C, Andrews G: Measuring personality disorder: the
use of self-report questionnaires. J Personal Disord
6:125133, 1992

Hyler SE: Personality Diagnostic Questionnaire-4 (PDQ-4).


New York, New York State Psychiatric Institute, 1994
Kaye A, Shea MT: Personality disorders, personality traits,
and defense mechanisms, in Handbook of Psychiatric
Measures. Edited by Pincus HA, Rush AJ, First MB, et
al. Washington, DC, American Psychiatric Association,
2000, pp 713749
Klein MH, Benjamin LS, Rosenfeld R, et al: The Wisconsin
Personality Disorders Inventory: development, reliability, and validity. J Personal Disord 7:285303, 1993
Leary T: Interpersonal Diagnosis of Personality: A Functional Theory and Methodology for Personality Evaluation. New York, Ronald Press, 1957
Lilienfield SO, Andrews BP: Development and preliminary
validation of a self-report measure of psychopathic personality traits in noncriminal populations. J Pers Assess
60:488524, 1996
Livesley WJ, Jackson DN: Dimensional Assessment of Personality ProblemsBasic Questionnaire. Port Huron,
MI, Sigma Assessment Systems (in press)
Loranger AW: International Personality Disorders Examination Manual. Odessa, FL, Psychological Assessment Resources, 1999
Loranger AW, Sartorius N, Andreoli A, et al: The International
Personality Disorder Examination: The World Health Organization/Alcohol, Drug Abuse, and Mental Health
Administration International Pilot Study of Personality
Disorders. Arch Gen Psychiatry 51:215224, 1994
McDermut W, Zimmerman M: The effect of personality disorders on outcome in the treatment of depression, in
Mood and Anxiety Disorders. Edited by Rush AJ. Philadelphia, PA, Current Science, 1997, pp 321338
Millon T, Davis R, Millon C: Manual for the MCMI-III. Minneapolis, MN, National Computer Systems, 1997
Morey LC: Personality Assessment Inventory Professional
Manual. Odessa, FL, Psychological Assessment Resources, 1991
Morey LC, Waugh MH, Blashfield RK: MMPI scales for
DSM-III personality disorders: their derivation and correlates. J Pers Assess 49:245251, 1985
Pfohl B, Blum N, Zimmerman M: Structured Interview for
DSM-IV Personality. Washington, DC, American Psychiatric Press, 1997
Piersma HL: The MCMI-II as a treatment outcome measure
for psychiatric inpatients. J Clin Psychol 45:8793, 1989
Pilkonis PA, Heape CL, Ruddy J, et al: Validity in the diagnosis of personality disorders: the use of the LEAD standard. Psychol Assess 3:4654, 1991
Raine A: The SPQ: a scale for the assessment of schizotypal
personality based on DSM-III-R criteria. Schizophr Bull
17:555564, 1991
Raine A, Benishay D: The SPQ-B: a brief screening instrument for schizotypal personality disorder. J Personal
Disord 9:346355, 1995
Raskin RN, Terry H: A principal-components analysis of the
narcissistic personality inventory and further evidence of
its construct validity. J Pers Soc Psychol 54:890902, 1988

Assessment Instruments and Standardized Evaluation

Regier DA, Myers JK, Kramer M, et al: The NIMH Epidemiologic Catchment Area (ECA) program: historical context, major objectives, and study population characteristics. Arch Gen Psychiatry 41:934941, 1984
Ricciardi JN, Baer L, Janice MA, et al: Changes in DSM-III-R
Axis II diagnoses following treatment of obsessivecompulsive disorder. Am J Psychiatry 149:829831, 1992
Riso LN, Klein DN, Anderson RL, et al: Concordance between patients and informants on the Personality Disorder Examination. Am J Psychiatry 151:568573, 1994
Robins LN, Helzer JE, Croughan J, et al: The National Institute of Mental Health Diagnostic Interview Schedule:
its history, characteristics, and validity. Arch Gen Psychiatry 42:381389, 1981
Robins LN, Wing J, Wittchen HU: The Composite International Diagnostic Interview: an epidemiologic Instrument suitable for use in conjunction with different diagnostic systems and in different cultures. Arch Gen
Psychiatry 12:10691077, 1988
Rogers R: Standardizing DSM-IV diagnoses: the clinical applications of structured interviews. J Pers Assess 81:220
225, 2003
Shea MT, Pilkonis PA, Beckham PA, et al: Personality disorders and treatment outcome in the NIMH Treatment of
Depression: Collaborative Research Program. Am J Psychiatry 147:711718, 1990
Sullivan HS: The Interpersonal Theory of Psychiatry. New
York, WW Norton, 1953
Trull TJ, Widiger TA: Structured Interview for the FiveFactor Model of Personality. Odessa, FL, Psychological
Assessment Resources, 1997
Tyrer P: Personality Disorders: Diagnosis, Management and
Course. London, Wright, 1988
Westen D: Divergences between clinical research methods
for assessing personality disorders: implications for research and the evolution of Axis II. Am J Psychiatry 154:
895903, 1997
Westen D, Shedler J: Refining the measurement of Axis II: a
Q-sort procedure for assessing personality pathology.
Assessment 5:333353, 1998
Widiger TA: The DSM-III-R categorical personality disorder
diagnoses: a critique and an alternative. Psychol Inq
4:7590, 1993

101

Widiger TA: Personality disorders, in Handbook of Assessment and Treatment Planning for Psychological Disorders. Edited by Antony MM, Barlow DH. New York,
Guilford, 2002, pp 453480
Widiger TA, Coker LA: Assessing personality disorders, in
Clinical Personality Assessment: Practical Approaches,
2nd Edition. Edited by Butcher JN. New York, Oxford
University Press, 2002, pp 380394
Widiger TA, Spitzer RL: Sex bias in the diagnosis of personality disorders: conceptual and methodological issues.
Clin Psychol Rev 11:122, 1991
Widiger TA, Mangine S, Corbitt EM, et al: Personality Disorder
InterviewIV: A Semi-Structured Interview for the Assessment of Personality Disorders, Professional Manual.
Odessa, FL, Psychological Assessment Resources, 1995
World Health Organization: The ICD-10 Classification of
Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva, Switzerland,
World Health Organization, 1992
Yang J, McCrae RR, Costa PT, et al: The cross-cultural generalizability of Axis-II constructs: an evaluation of two personality disorder assessment instruments in the Peoples
Republic of China. J Personal Disord 14:249263, 2000
Zanarini MC, Gunderson JG, Frankenburg FR, et al: The Revised Diagnostic Interview for Borderlines: discriminating borderline personality disorder from other Axis II
disorders. J Personal Disord 3:1018, 1989
Zanarini MC, Frankenburg FR, Sickel AE, et al: Diagnostic
Interview for DSM-IV Personality Disorders. Boston,
MA, Laboratory for the Study of Adult Development,
McLean Hospital, and the Department of Psychiatry,
Harvard University, 1996
Zimmerman M: Diagnosing personality disorders: a review
of issues and research methods. Arch Gen Psychiatry
51:225245, 1994
Zimmerman M, Mattia JI: Differences between clinical and
research practices in diagnosing borderline personality
disorder. Am J Psychiatry 156:15701574, 1999a
Zimmerman M, Mattia JI: Psychiatric diagnosis in clinical
practice: is comorbidity being missed? Compr Psychiatry 40:182191, 1999b
Zimmerman M, Coryell W, Black DW: A method to detect intercenter differences in the application of contemporary
diagnostic criteria. J Nerv Ment Dis 181:130134, 1993

This page intentionally left blank

6
Course and Outcome of
Personality Disorders
Carlos M. Grilo, Ph.D.
Thomas H. McGlashan, M.D.

A personality disorder is defined in DSM-IV-TR (American Psychiatric Association 2000) as an enduring pattern of inner experience and behavior that deviates
markedly from the expectations of the individuals culture, is pervasive and inflexible, has an onset in adolescence or early adulthood, is stable over time, and leads
to distress or impairment (p. 685). The diagnostic construct of personality disorder has evolved considerably
over the past few decades (see Skodol 1997 for a detailed
ontogeny of the DSM system; and see Chapter 1, Personality Disorders: Recent History and Future Directions, for a historical overview). Substantial changes
have occurred in both the number and types of specific
personality disorder diagnoses over time, as well as in
the admixture of criteria (Sanislow and McGlashan
1998) representing possible manifestations of personality disorders (i.e., DSM-IV-TR specifies that the enduring pattern can be manifested by problems in at least
two of the following areas: cognition, affectivity, interpersonal functioning, or impulse control). One central tenetthat a personality disorder reflects a persistent, pervasive, enduring, and stable patternhas not

changed. The concept of stability is salient in both major


classification systems, DSM-IV-TR and ICD-10 (World
Health Organization 1992), although the two systems
differ somewhat in their classification and definitions
for personality disorders and thus demonstrate only
moderate convergence for some diagnoses (Ottosson et
al. 2002). The extent of stability of personality disorders
remains uncertain (Shea and Yen 2003; Tyrer and Simonsen 2003). This chapter provides an overview of the
course and outcome of personality disorders and synthesizes the empirical literature on the stability of personality disorders.

STABILITY AS THE CENTRAL TENET


PERSONALITY DISORDERS

OF

The concept of stability has remained a central tenet of


personality disorders throughout the various editions
of DSM, dating back to the first edition, published in
1952. In what some experts have referred to as a bold
103

104

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

step (Tyrer and Simonsen 2003), personality disorders were placed on a separate Axis (Axis II) of the
multiaxial DSM-III (American Psychiatric Association
1980), published in 1980. DSM-III stated that the separation to Axis II was intended, in part, to encourage
clinicians to assess the possible presence of disorders
that are frequently overlooked when attention is
directed to the usually more florid Axis I disorder.
Conceptually, this separation reflected the putative
stability of personality disorders relative to the episodically unstable course of Axis I psychiatric disorders (Grilo et al. 1998; Skodol 1997).

FIRST- AND SECOND-GENERATION


RESEARCH STUDIES ON STABILITY
First, we provide a brief review of the empirical literature through the end of the twentieth century. This period can be thought of as including the first generation
(mostly clinical-descriptive accounts) and the second
generation (the emerging findings based on attempts at
greater standardization of diagnoses and assessment
methods) of research efforts on personality disorders.
Second, we provide a brief overview of methodological
problems and conceptual gaps that characterize this literature and that must be considered when interpreting
ongoing research and designing future studies. Third,
we summarize emerging findings from ongoing longitudinal studies that have shed light on a number of key
issues about the course of personality disorders.

OVERVIEW OF THE LITERATURE


THROUGH 1999
A number of previous reviews have been published
addressing aspects of the course and outcome of
personality disorders (Grilo and McGlashan 1999;
Grilo et al. 1998; McDavid and Pilkonis 1996; Perry
1993; Ruegg and Frances 1995; Stone 1993; Zimmerman 1994). These reviews, although varied, have
agreed on the pervasiveness of methodological problems that characterize much of the literature and
thereby preclude any firm conclusions regarding the
nature of the stability of personality disorders. The reviews, however, have also generally agreed that available research raises questions regarding many aspects
of the construct validity of personality disorders (Zimmerman 1994), including their hypothesized high degree of stability (Grilo and McGlashan 1999).

The few early (pre-DSM-III era) studies of the


course of personality disorders reported findings that
borderline (Carpenter and Gunderson 1977; Grinker
et al. 1968) and antisocial (Maddocks 1970; Robins et
al. 1977) personality disorders were highly stable. Carpenter and Gunderson (1977), for example, reported
that the impairment in functioning observed for borderline personality disorder (BPD) was comparable
with that observed for patients with schizophrenia
over a 5-year period. As previously noted (Grilo et al.
1998), the dominant clinical approach to assessing personality disorder diagnoses based partly on treatment
refractoriness naturally raises the question of whether
these findings simply reflect a tautology.
The separation of personality disorders to Axis II in
DSM-III contributed to increased research attention to
these clinical problems (Blashfield and McElroy 1987).
The development and utilization of a number of structured and standardized approaches to clinical interviewing and diagnosis during the 1980s represented
notable advances (Zimmerman 1994). The greater attention paid to defining the criteria required for diagnosis in the classification systems and by researchers
during the development of standardized interviews
greatly facilitated research efforts in this field.
In our previous reviews of the DSM-III and DSMIII-R (American Psychiatric Association 1987) studies,
we concluded that the available research suggested that
personality disorders demonstrate only moderate stability and that, although personality disorders are generally associated with negative outcomes, they can
improve over time and can benefit from specific treatments (Grilo and McGlashan 1999, p. 157). In our 1998
review (Grilo et al. 1998), we noted that the 20 selected
studies of DSM-III-R criteria generally found low to
moderate stability of any personality disorder over relatively short follow-up periods (6 to 24 months). For example, the major studies that employed diagnostic interviews reported kappa coefficients for the presence of
any personality disorder of 0.32 (Johnson et al. 1997),
0.40 (Ferro et al. 1998), 0.50 (Loranger et al. 1994), and
0.55 (Loranger et al. 1991). Especially noteworthy is that
the stability coefficients for specific personality disorder diagnoses (in the few cases in which they could be
calculated given the sample sizes) were generally
lower. In addition, follow-up studies of adolescents
diagnosed with personality disorders also reported
modest stability; for example, Mattanah et al. (1995)
reported a 50% rate of stability for any personality disorder at 2-year follow-up. More recently, Grilo and colleagues (2001) also found modest stability in dimensional personality disorder scores in this adolescent

Course and Outcome of Personality Disorders

follow-up study. Squires-Wheeler et al. (1992), as part of


the New York State High-Risk Offspring Study, reported low stability for schizotypal personality disorder and features, although the stability was higher for
the offspring of patients with schizophrenia than for
those with mood disorders or control subjects.
Subsequently, we (Grilo and McGlashan 1999) reviewed nine reports of longitudinal findings for personality disorder diagnoses published in 1997 and 1998.
In terms of specific diagnoses, the studies generally reported moderate stability (kappa approximately 0.5)
for BPD and antisocial personality disorder (ASPD).
These reports, like most of the previous literature, had
small sample sizes and infrequently followed more
than one personality disorder.

CONCEPTUAL AND METHODOLOGICAL


QUESTIONS ABOUT COURSE
Previous reviews of personality disorders have raised
many methodological problems. Common limitations
highlighted include small sample sizes; concerns about
nonstandardized assessments, interrater reliability,
blindness to baseline characteristics, and narrow assessments; failure to consider alternative (e.g., dimensional) models of personality disorder; reliance on only
two assessments typically over short follow-up periods; insufficient attention to the nature and effects of
co-occurring Axis I and Axis II diagnoses; and inattention to treatment effects. Diagnoses other than ASPD
and BPD have received little attention. Particularly
striking is the absence of relevant comparison or
control groups in the longitudinal literature. We comment briefly on a few of these issues.

Reliability
Reliability of assessments represents a central issue for
any study of course and outcome. The creation of standardized instruments for collecting data was a major
development of the 1980s (Loranger et al. 1991; Zimmerman 1994). Such instruments, however, were lessthan-perfect assessment methods and have been criticized for a variety of reasons (Westen 1997; Westen
and Shedler 1999). It is critical to keep in mind that interrater reliability and testretest reliability represent
the limits (or ceiling) for estimating the stability of a
construct.
Previous reviews (Grilo and McGlashan 1999; Zanarini et al. 2000; Zimmerman 1994) of reliabilities for
Axis II diagnostic interviews have generally reported

105

median interrater reliabilities of roughly 0.70 and


short-interval testretest reliabilities of 0.50 for diagnoses. These reliabilities compare favorably with those
generally reported for diagnostic instruments for
Axis I psychiatric disorders. Both interrater and test
retest reliability coefficients tend to be higher for dimensional scores than for categorical diagnoses of personality disorders. Another finding of note is that even
when experts administer diagnostic interviews, the degree of convergence or agreement produced by two
different interviews administered only 1 week apart is
limited (Oldham et al. 1992).

Reliability and Change


Testretest reliability is also relevant for addressing, in
part, the well-known problem of regression to the
mean in repeated measures studies (Nesselroade et
al. 1980). It has been argued that the multiwave or repeated measures approach lessens the effects of regression to the mean (Lenzenweger 1999). This argument may be true in terms of the obvious decreases in
severity with time (i.e., very symptomatic participants
meeting eligibility at study entry are likely to show
some improvement because, by definition, they are already reporting high levels of symptoms). However,
other effects need to be considered whenever assessments are repeated within a study. As cogently noted
by Shea and Yen (2003), repeated measures studies of
both Axis II (Loranger et al. 1991) and Axis I (Robins
1985) disorders have found hints that participants systematically report or endorse fewer problems during
repeated interviews to reduce interview time. For example, Loranger et al. (1991), in his testretest study of
the Personality Disorder Examination interview (Loranger 1988) conducted between 1 and 26 weeks after
baseline, documented significant decreases in personality disorder criteria for all but two of the DSM-III-R
diagnoses. Recall that the Personality Disorder Examination, which requires skilled and trained research
clinicians, has a required minimum duration stipulation of 5 years for determining persistence and pervasiveness of the criteria being assessed. Thus, the magnitude of changes observed during such a short period
of time, which was shown to be unrelated to statetrait effects, reflects some combination of the following: regression to the mean, error in either or both the
baseline and repeated assessments, and overreporting
by patients at hospital admission and underreporting
during retest at discharge (Loranger et al. 1991; Shea
and Yen 2003). These phenomena were discussed further by Gunderson and colleagues (2000).

106

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Categorical Versus Dimensional Approaches


Long-standing debate regarding the conceptual and
empirical advantages to dimensional models of
personality disorders (Frances 1982; Livesley et al.
1992; Loranger et al. 1994; Widiger 1992) has accompanied the DSM categorical classification system. In
Chapter 3, Categorical and Dimensional Models of
Personality Disorders, Widiger and Mullins-Sweatt
address this issue. We comment only briefly on the literature that applies specifically to the issue of course
of personality disorders. Overall, longitudinal studies
of personality disorder have reported moderate levels
of stability for dimensional scores for most personality
disorders, with the stability coefficients tending to be
higher than for categorical or diagnostic stability
(Ferro et al. 1998; Johnson et al. 1997; Klein and Shih
1998; Loranger et al. 1991, 1994).

Comorbidity
Most studies have ascertained participants who meet
criteria for multiple Axis I and Axis II diagnoses. This
problem of diagnostic overlap, or comorbidity, represents a well-known, long-standing major challenge
(Berkson 1946) in working with clinical samples. One
expert and critic of DSM (Tyrer 2001), in speaking of the
spectre of comorbidity, noted that the main reason
for abandoning the present classification is summed up
in one word, comorbidity. Comorbidity is the nosologists nightmare; it shouts, you have failed (p. 82). We
suggest, however, that such clinical realities (multiple
presenting problems that are especially characteristic of
treatment-seeking patients) represent not only potential confounds but also potential opportunities to understand personality and dysfunctions of personality
better. Comorbidity begs the question: what are the
fundamental personality dimensions and disorders of
personality, and how do their courses influence (and
conversely, how are their courses affected by) the presence and course of Axis I psychiatric disorders?

Continuity
A related issue pertaining to course concerns longitudinal comorbidities (Kendell and Clarkin 1992) or
continuities. An obvious example is that conduct
disorder during adolescence is required for the diagnosis of ASPD to be given to adults. This definitional
isomorphism is one likely reason for the consistently
strong associations between conduct disorder and
later ASPD in the literature. This association is, how-

ever, more than an artifactual relationship, because


longitudinal research has clearly documented that
children and adolescents with behavior disorders
have substantially elevated risk for antisocial behavior during adulthood (Robins 1966). More generally,
studies with diverse recruitment and ascertainment
methods reported that disruptive behavior disorders
during the adolescent years prospectively predicted
personality disorders during young adulthood (Bernstein et al. 1996; Lewinsohn et al. 1997; Myers et al.
1998; Rey et al. 1995). The Yale Psychiatric Institute
follow-up study found that personality disorder diagnoses in adolescent inpatients prospectively predicted greater drug use problems but not global functioning (Levy et al. 1999).
The importance of considering comorbidity is underscored in the findings of the longitudinal study by
Lewinsohn et al. (1997). They found that the apparent
longitudinal continuity noted for disruptive behavioral
disorders during adolescence and subsequent ASPD in
adulthood was accounted for, in part, by Axis I psychiatric comorbidity. A longitudinal study of young adult
men found that personality disorders predicted the
subsequent onset of psychiatric disorders during a
2-year follow-up, even after controlling for previous
psychiatric history (Johnson et al. 1997).

Comorbidity and Continuity Models


A variation of the comorbidity concept is that certain
disorders may be associated with one another in a number of possible ways over time. A variety of models
have been proposed for the possible relationships between Axis II and Axis I disorders (Dolan-Sewell et al.
2001; Lyons et al. 1997; Tyrer et al. 1997). These include,
for example, the predisposition or vulnerability model,
the complication or scar model, the pathoplasty or exacerbation model, and various spectrum models. We
emphasize that these models do not necessarily assume
categorical entities. Indeed, an especially influential
spectrum model proposed by Siever and Davis (1991)
posits four psychobiological dimensions to account for
Axis II and Axis I psychopathology. The Cloninger et
al. (1993) psychobiological model of temperament and
character represents another valuable approach that
considers dimensions across personality and psychopathology. More broadly, Krueger (Krueger 1999;
Krueger and Tackett 2003) noted that although most research has focused on pairs of constructs (i.e., Axis II
and Axis I associations), it seems important to examine
the multivariate structure of the personality-psychopathology domain (p. 109).

Course and Outcome of Personality Disorders

107

Age (Early Onset)

Age and the Aging Process

A related point, stressed by Widiger (2003), is that personality disorders need to be more clearly conceptualized and carefully characterized as having an early onset. However, the validity of personality disorders in
adolescents remains controversial (Krueger and Carlson 2001). It can be argued, for example, that determining early onset of personality disorders is impossible because adolescence is a period of profound
changes and flux in personality and identity. A recent
critical review of the longitudinal literature on personality traits throughout the lifespan revealed that personality traits are less stable during childhood and adolescence than they are th roughout adulth ood
(Roberts and DelVecchio 2000). Roberts and DelVecchios (2000) meta-analysis of data from 152 longitudinal studies of personality traits revealed that rankorder consistency for personality traits increased
steadily throughout the lifespan; testretest correlations (over 6.7-year time intervals) increased from 0.31
(during childhood) to 0.54 (during college), to 0.64
(age 30), to a high of 0.74 (ages 5070).
Nonetheless, if childhood precursors of personality disorders could be identified (as in the case of conduct disorder for ASPD), they could become part of
the diagnostic criteria and thus create some degree of
longitudinal continuity in the diagnostic system. Myers et al. (1998), for example, found that early onset
(before 10 years of age) of conduct disorder problems
predicted subsequent ASPD. More generally, temperamental vulnerabilities or precursors to personality
disorders have been posited as central in a variety of
models of personality disorders (Cloninger et al. 1993;
Siever and Davis 1991). Specific temperamental features evident in childhood have been noted to be precursors for diverse personality disorders (Paris 2003;
Rettew et al. 2003; Wolff et al. 1991) as well as for differences in interpersonal functioning (Newman et al.
1997) in adulthood. For example, studies have noted
early odd and withdrawn patterns for schizotypal
personality disorder in adults (Wolff et al. 1991) and
shyness for avoidant personality disorder (Rettew et
al. 2003). Speaking more generally, although the
degree of stability for personality traits is higher
throughout adulthood than throughout childhood
and adolescence (Roberts and DelVecchio 2000), longitudinal analyses of personality data have revealed
that the transition from adolescence to adulthood is
characterized by greater personality continuity than
change (Roberts et al. 2001).

Another age issue concerns the aging process itself.


Considerable research suggests that personality remains relatively stable thorough adulthood (Heatherton and Weinberger 1994; Roberts and DelVecchio
2000) and is highly stable after age 50 (Roberts and
DelVecchio 2000). Little is known, however, about personality disorders in older persons (Abrams et al. 1998).
The recent 12-year follow-up of personality disorders
as part of the Nottingham Study of Neurotic Disorders
(Seivewright et al. 2002) documented substantial
changes in personality disorder trait scores based on
blind administration of a semistructured interview.
Seivewright and colleagues (2002) reported that Cluster
B personality disorder diagnoses (ASPD, histrionic)
showed significant improvements, whereas Cluster A
and Cluster C diagnoses appeared to worsen with age.
Although the Seivewright et al. (2002) findings are limited somewhat by the two-point cross-sectional assessment (little is known about the intervening period), Tyrer and colleagues (1983) previously reported good
reliability (weighted kappa of 0.64) for this diagnostic
interview over a 3-year testretest period. These findings echo somewhat the results of the seminal Chestnut
Lodge follow-up studies (McGlashan 1986a, 1986b) that
suggested distinctions between BPD and schizotypal
personality disorders, decreases in impulsivity and interpersonal instability with age, and increased avoidance with age. There are other reports of diminished
impulsivity with increasing age in BPD (Paris and
Zweig-Frank 2001; Stevensen et al. 2003), although this
type of reduction was not observed in a recent prospective analysis of individual BPD criteria (McGlashan et
al. in press).
The reader is referred to Judd and McGlashan (2003)
for detailed accounts of four specific cases that elucidate the course and outcome of BPD. These detailed
case studies, based on rich clinical material available
through the Chestnut Lodge Study, demonstrate the
considerable heterogeneity in the course of BPD.

Summary and Implications


To resolve these complex issues, complementary research efforts are required, with large samples of both
clinical and community populations. It is clear that
prospective longitudinal studies with repeated assessments over time are needed to understand the course
of personality disorders. Such studies must consider
(and cut across) different developmental eras, broad
domains of functioning, and multimodal approaches
to personality and disorders of personality. These ap-

108

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

proaches have, in fact, been performed with personality traits (Roberts et al. 2001) and with other forms of
psychiatric problems and have yielded invaluable insights. Notable are the contributions of the National
Institutes of Health (NIH)funded multisite efforts on
depression (Collaborative Depression Study; Katz et
al. 1979) and anxiety (Harvard/Brown Anxiety Research Program; Keller 1991).

REVIEW OF RECENT EMPIRICAL ADVANCES


AND UNDERSTANDING OF STABILITY
Of particular relevance for this review are three prospective studies on the longitudinal course of adult
personality disorders funded by the NIH during the
1990s. These studies included the Longitudinal Study
of Personality Disorders (Lenzenweger 1999), the
McLean Study of Adult Development (Zanarini et al.
2003), and the multisite Collaborative Longitudinal
Personality Disorders Study (CLPS; Gunderson et al.
2000). NIH also funded a community-based prospective longitudinal study of personality, psychopathology, and functioning of children/adolescents and their
mothers (Children in the Community Study; Brook et
al. 2002) that began in 1983. These four studies are especially noteworthy in that they, to varying degrees,
partly correct for a number of the conceptual and
methodological issues noted earlier. These studies utilized multiple and standardized assessment methods,
carefully considered training and reliability, andperhaps most notablymultiwave repeated assessments
that are essential for determining longitudinal change.
They have employed, to varying degrees, multiple assessment methods and have considered personality
and its disorders (personality disorders) as well as
Axis I psychiatric disorders. Collectively, these studies have provided valuable insights into the complexities of personality (traits and disorders) and its vicissitudes over time.

Longitudinal Study of Personality Disorders


The Longitudinal Study of Personality Disorders
(Lenzenweger 1999; Lenzenweger et al. 1997) assessed 250 participants drawn from Cornell University at three points over a 4-year period. It utilized a
semistructured diagnostic interview (International
Personality Disorders Examination; Loranger et al.
1994) and a self-report measure (Millon Clinical Multiaxial InventoryII) to obtain complementary infor-

mation on personality. Of the 250 participants, 129


met criteria for at least one personality disorder and
121 did not meet any personality disorder diagnosis.
Dimensional scores for the personality disorders
were characterized by significant levels of stability
on both the interview and self-report measures. Stability coefficients for the total number of personality
disorder features ranged from 0.61 to 0.70. Cluster B
personality disorders had the highest stability coefficients and Cluster A personality disorders had the
lowest. Personality disorder dimensions showed significant declines over time, and the decline was more
rapid for the personality disorder group than for the
nondisordered group. Axis I psychiatric disorders
(diagnosed in 63% of personality disorder subjects
and 26% of nonpersonality disorder subjects) did
not significantly influence changes in personality
disorder dimensions over time.
The Longitudinal Study of Personality Disorders
(Lenzenweger 1999) BPD findings are generally consistent (although the three-point assessment is an
important incremental contribution) with those previously reported by Trull and colleagues (1997, 1998)
in a prospective study of BPD features using two different assessment instruments administered to a college student sample assessed twice over a 2-year period.
The Longitudinal Study of Personality Disorders
(Lenzenweger 1999), however, is limited by its relatively homogeneous study group of college students,
its narrow developmental time frame, and most importantly the insufficient frequency of any personality
disorder diagnosis at a categorical (diagnostic) level to
allow analysis of a clinical entity. Lenzenweger (1999)
noted the need for repeated-measure longitudinal
data from clinically based personality disorder samples to address the question of the course and stability
of dysfunctions of personality.

McLean Study of Adult Development


The McLean Study of Adult Development (Zanarini
et al. 2003) is an ongoing prospective, longitudinal
study comparing the course and outcome of hospitalized patients with BPD with those of patients with
other personality disorders. It utilizes repeated assessments performed every 2 years (Zanarini et al.
2003). Zanarini et al. (2003) assessed personality disorders in 362 inpatients (290 with BPD and 72 with
other personality disorders) using two semistructured diagnostic interviews and administered assessments to characterize Axis I psychiatric disorders,

Course and Outcome of Personality Disorders

psychosocial functioning domains, and treatment


utilization. Of the patients diagnosed with BPD, remission was observed for 35% by year 2, 49% by year
4, and 74% by year 6. Recurrences were rare and were
reported for only 6% of those patients who achieved
a remission. The authors concluded that symptomatic improvement is both common and stable, even
among the most disturbed borderline patients, and
that the symptomatic prognosis for most, but not all,
severely ill borderline patients is better than previously recognized (p. 274).

Collaborative Longitudinal
Personality Disorders Study
The Collaborative Longitudinal Personality Disorders
Study (CLPS; Gunderson et al. 2000; McGlashan et al.
2000) is an ongoing prospective, longitudinal, repeated measures study designed to examine the
course and outcome of patients meeting DSM-IV
(American Psychiatric Association 1994) criteria for
one of four personality disorders: schizotypal, borderline, avoidant, and obsessive-compulsive. CLPS includes a comparison group of patients with major depressive disorder (MDD) without any personality
disorder. This comparison group was selected because of its episodic and fluctuating course (thought
to distinguish Axis I from Axis II) and because MDD
has been carefully studied in similar longitudinal designs (e.g., Collaborative Depression Study; Katz et al.
1979; Solomon et al. 1997). CLPS has employed multimodal assessments (Gunderson et al. 2000; Zanarini et
al. 2000) to prospectively follow and capture different
aspects of the fluctuating nature of personality disorders and dimensions (both interviewer-based and
self-report representing different conceptual models),
Axis I psychiatric disorders and symptoms, various
domains of psychosocial functioning, and treatment
utilization.
To date, the CLPS has reported on different concepts of categorical and dimensional stability of four
personality disorders over 12 months (Shea et al. 2002)
and 24 months (Grilo et al. 2004) using prospective
data obtained for 668 patients recruited from diverse
settings at four universities. Based on the traditional
testretest approach, blind repeated administration of
a semistructured interview conducted 24 months after
baseline revealed remission rates (based solely on
falling below DSM-IV diagnostic thresholds) ranging
from 50% (avoidant personality disorder) to 61%
(schizotypal personality disorder). Grilo et al. (2004)
applied lifetable survival analyses to prospective data

109

obtained using an assessment methodology modeled


after the Collaborative Depression Study (Keller et al.
1982) and the Longitudinal Interval Follow-Up Evaluation (Keller et al. 1987) methodology. These findings
are summarized in Figures 61 and 62.
Figure 61 shows the times to remission for the
four personality disorder groups and for the MDD
comparison group, which were calculated based on
parallel definitions of two consecutive months with
minimal symptoms (Grilo et al. 2004). As can be seen,
the MDD group had a significantly higher remission
rate than the personality disorder groups. This study
represents the first empirical demonstration of the
central tenet that personality disorders are characterized by greater degree of stability than the hypothesized episodic course of Axis I psychiatric disorders
(Grilo et al. 1998; Shea and Yen 2003.
The reader is referred to Shea and Yen (2003) for a
broader discussion of this issue. These researchers,
who have played roles in the CLPS as well as the longitudinal studies of depression (Collaborative Depression Study) and anxiety (Harvard/Brown Anxiety Research Program), provide an overview of the central
findings that pertain to the issue of stability as a distinction between Axis II and Axis I diagnoses (Shea
and Yen 2003). Briefly, comparison across the studies
(which can be done given the parallel assessment
instrumentation) reveals that personality disorders
demonstrate greater stability than Axis I mood and
anxiety disorders (as hypothesized) but show less diagnostic (categorical) stability than conceptualized.
Perhaps noteworthy is that the longitudinal studies
for both mood and anxiety disorders documented
much greater chronicity (much lower remission rates)
than previously known.
Returning to the CLPS findings (Grilo et al. 2004),
Figure 61 reveals that although personality disorders
were more stable than MDD, a substantial number of
remissions occurred during the 24 months of follow-up. Using the arbitrarily selected 2-month definition (2 months with two or fewer criteria) adopted
from the MDD field (Keller et al. 1982; Solomon et al.
1997), remission rates range from 33% (schizotypal
personality disorder) to 55% (obsessive-compulsive
personality disorder). Figure 62 shows the comparable remission rates if a very stringent definition of
12 consecutive months with two or fewer criteria is
adopted. As can be seen, the remission rates using the
12-month definition range from 23% (schizotypal personality disorder) to 38% (obsessive-compulsive personality disorder). Grilo et al. (2004) concluded that
these four personality disorders show substantial im-

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Proportion Not Remitted

110

Time From Intake (Months)

Figure 61.

Time to remission using a 2-month criterion.

AVPD=avoidant personality disorder; BPD =borderline personality disorder; MDD =major depressive disorder; OCPD=obsessivecompulsive personality disorder; STPD =schizotypal personality disorder.
Source. Grilo et al. 2004. Copyright 2004 by the American Psychological Association. Reprinted with permission.

provements in symptomatology over a 2-year period


even when a stringent definition is used.
The CLPS also provided complementary analyses
using dimensional approaches for 12-month (Shea et
al. 2002) and 24-month (Grilo et al. 2004) follow-ups.
Grilo et al. (2004) documented a significant decrease in
the mean proportion of criteria met in each of the personality disorder groups over time, which is suggestive of decreased severity. However, when the relative
stability of individual differences was examined across
the multiwave assessments (baseline and 6-, 12-, and
24-month time points), a high level of consistency was
observed as evidenced by correlation coefficients ranging from 0.53 to 0.67 for proportion of criteria met between baseline and 24 months. Grilo and colleagues
(2004) concluded that patients with personality disorder are consistent in terms of their rank order of personality disorder criteria (i.e., that individual differences in personality disorder features are stable),
although they may fluctuate in the severity or number
of personality disorder features over time. It is worth
noting that the range of the stability coefficients was

quite similar to that documented by the Longitudinal


Study of Personality Disorders (Lenzenweger 1999) for
a nonclinical sample.
In contrast to their symptomatic improvement,
however, patients with personality disorders show
less significant and more gradual improvement in
their functioning, particularly in social relationships
(Skodol et al. in press). In addition, depressed patients
with personality disorders show longer time to remission from major depressive disorder (Grilo et al. in
press). Because personality psychopathology usually
begins in adolescence or early adulthood, the potential
for delays in occupational and interpersonal development is greatand even after symptomatic improvement, it might take time to overcome deficits and
make up the necessary ground to achieve normal
functioning. Developmental issues for patients with
personality disorders are discussed in more detail in
Chapter 11, Developmental Issues.
Several recent reports from the CLPS are also relevant here given the issue of longitudinal comorbidities
and continuities. Shea and colleagues (2004) examined

111

Proportion Not Remitted

Course and Outcome of Personality Disorders

Time From Intake (Months)

Figure 62.

Time to remission using a 12-month criterion.

AVPD=avoidant personality disorder; BPD=borderline personality disorder; OCPD =obsessive-compulsive personality disorder;
STPD=schizotypal personality disorder.
Source. Grilo et al. 2004. Copyright 2004 by the American Psychological Association. Reprinted with permission.

the time-varying (longitudinal) associations between


personality disorders and psychiatric disorders, in part
guided by the Siever and Davis (1991) cross-cutting
psychobiological dimension model. BPD demonstrated
significant associations with certain psychiatric disorders (MDD and posttraumatic stress disorder), whereas
avoidant personality disorder was significantly associated with two anxiety disorders (social phobia and obsessive-compulsive disorder). While these findings
were consistent with predictions based on the Siever
and Davis (1991) model, other personality disorders
(schizotypal and obsessive-compulsive) did not demonstrate significant longitudinal associations. Gunderson et al. (2004) followed up on the Shea et al. (2004)
findings regarding changes in BPD and MDD by performing a more fine-grained analysis of specific
changes in the two disorders using 3 years of longitudinal data. Changes (improvements) in BPD severity preceded improvements in MDD but not vice versa (Gunderson et al. 2004).
Another recent report (Warner et al. 2004) examined whether personality traits are stable in patients
with personality disorders and tested the hypothesis

that the stability of these personality disorders is due


in part to the stability in these traits (Lyman and Widiger 2001). A series of latent longitudinal models
tests whether changes in specific traits prospectively
predicted changes in relevant personality disorders.
Warner et al. (2004) documented significant crosslagged relationships between changes in specific
traits and subsequent (later) changes for schizotypal,
borderline, and avoidant personality disorders but
not for obsessive-compulsive personality disorder.
McGlashan and colleagues (in press) examined the
individual criteria for schizotypal, borderline, avoidant, and obsessive-compulsive personality disorders
and how they changed over a 2-year period. The individual criteria for these four personality disorders
showed varied patterns of stability and change over
time. Overall, within personality disorders, the relatively fixed (least changeable) criteria were generally
more traitlike (and attitudinal), whereas the more fluctuating criteria were generally behavioral (or reactive). McGlashan and colleagues (in press) posited
that perhaps personality disorders are hybrids of traits
and symptomatic behaviors and that it is the interac-

112

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

tion of these over time that helps to define the observable diagnostic stability. Collectively, along with the
recent CLPS efforts, these findings suggest that personality disorder traits are stable over time and across
developmental eras and may generate intra- and interpersonal conflicts that result in behaviors symptomatic of personality disorders (which are less stable over
time).

Children in the Community Study


The Children in the Community Study (D.W. Brook et
al. 2002; J.S. Brook et al. 1995) is an especially impressive longitudinal effort that has already provided a
wealth of information about the course of personality
and behavioral traits, psychiatric problems, substance
abuse, and adversities. It is an ongoing prospective
study of nearly 1,000 families with children aged 1 to 10
years originally recruited in1975 in New York State
using a random sampling procedure. The study has
employed repeated multimodal assessments and has
followed over 700 participants since childhood and
through the development eras of childhood, adolescence, and early adulthood. This landmark study has
provided data that speaks to the critical issues of longitudinal comorbidities and continuities. In a series of papers, the collaborating researchers have documented
important findings relevant to the issues raised in this
review but especially to the critical issues of continuity
of risk and functioning across developmental eras.
These include documentation of the validity of certain
forms of dramatic-erratic personality disorders in adolescents (Crawford et al. 2001a, 2001b); findings of agerelated changes in personality disorder traits, including
their moderate levels of stability throughout adolescence and early adulthood (Johnson et al. 2000b); and
indications that early forms of behavioral disturbances
predict personality disorders in adolescents and that
personality disorders during adolescence, in addition
to demonstrating significant levels of continuity into
adulthood, also predict psychiatric disorders, suicidality, and violent and criminal behavior during young
adulthood (Johnson et al. 2000a, 2000b). Collectively,
these findings support the continuity and persistence
of personality disturbances, although their development pathways are not yet understood.

SUMMARY
We have reviewed the literature regarding the course
and stability of personality disorders. We once again

conclude that personality disorders demonstrate only


moderate stability and that they can improve over time.
This conclusion is offered with less caution than during
our previous reviews (Grilo and McGlashan 1999),
given some notable advances in research. We also conclude that when personality disorders are considered
dimensionally, the degree of stability is substantial.
Emerging work has suggested that personality disorder traits, although deviant, are stable over time and
across developmental eras and may generate intra- and
interpersonal conflicts that result in personality disordersymptomatic behaviors (which are less stable over
time). Future research in personality disorders is necessary to dissect and understand this trait/state interaction and track its vicissitudes across time and circumstances.

REFERENCES
Abrams RC, Spielman LA, Alexopoulos GS, et al: Personality disorder symptoms and functioning in elderly
depressed patients. Am J Geriatr Psychiatry 6:2430,
1998
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition. Washington,
DC, American Psychiatric Association, 1980
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition, Revised. Washington, DC, American Psychiatric Association, 1987
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition. Washington,
DC, American Psychiatric Association, 1994
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association,
2000
Berkson J: Limitations of the application of fourfold table
analysis to hospital data. Biometrics Bulletin 2:4753,
1946
Bernstein DP, Cohen P, Skodol AE, et al: Childhood antecedents of adolescent personality disorders. Am J Psychiatry 153:907913, 1996
Blashfield RK, McElroy RA: The 1985 journal literature on
the personality disorders. Compr Psychiatry 28:536
546, 1987
Brook DW, Brook JS, Zhang C, et al: Drug use and the risk of
major depressive disorder, alcohol dependence, and
substance use disorders. Arch Gen Psychiatry 59:1039
1044, 2002
Brook JS, Whiteman M, Cohen P, et al: Longitudinally predicting late adolescent and young adult drug use: childhood and adolescent precursors. J Am Acad Child Adolesc Psychiatry 34:12301238, 1995

Course and Outcome of Personality Disorders

Carpenter WT, Gunderson JG: Five year follow-up comparison of borderline and schizophrenic patients. Compr
Psychiatry 18:567571, 1977
Cloninger CR, Svrakic DM, Przybeck TR: A psychobiological model of temperament and character. Arch Gen Psychiatry 50:975990, 1993
Crawford TN, Cohen P, Brook JS: Dramatic-erratic personality
disorder symptoms, I: continuity from early adolescence
into adulthood. J Personal Disord 15:319335, 2001a
Crawford TN, Cohen P, Brook JS: Dramatic-erratic personality disorder symptoms, II: developmental pathways
from early adolescence to adulthood. J Personal Disord
15:336350, 2001b
Dolan-Sewell RT, Krueger RF, Shea MT: Co-occurrence with
syndrome disorders, in Handbook of Personality Disorders. Edited by Livesley WJ. New York, Guilford, 2001,
pp 84104
Ferro T, Klein DN, Schwartz JE, et al: 30-month stability of
personality disorder diagnoses in depressed outpatients. Am J Psychiatry 155:653659, 1998
Frances A: Categorical and dimensional systems of personality diagnosis: a comparison. Compr Psychiatry 23:516
527, 1982
Grilo CM, McGlashan TH: Stability and course of personality disorders. Curr Opin Psychiatry 12:157162, 1999
Grilo CM, McGlashan TH, Oldham JM: Course and stability
of personality disorders. J Pract Psychiatry Behav Health
4:6175, 1998
Grilo CM, Becker DF, Edell WS, et al: Stability and change of
personality disorder dimensions in adolescents followed up two years after psychiatric hospitalization.
Compr Psychiatry 42:364368, 2001
Grilo CM, Shea MT, Sanislow CA, et al: Two-year stability
and change in schizotypal, borderline, avoidant, and
obsessive-compulsive personality disorders. J Consult
Clin Psychol 72:767775, 2004
Grilo CM, Sanislow CA, Shea MT, et al: Two-year prospective naturalistic study of remission from MDD as a
function of personality disorder comorbidity. J Consult
Clin Psychol (in press)
Grinker RR, Werble B, Drye RC: The Borderline Syndrome.
New York, Basic Books, 1968
Gunderson JG, Shea MT, Skodol AE, et al: The Collaborative
Longitudinal Personality Disorders Study, I: development, aims, designs and sample characteristics. J Personal Disord 14:300315, 2000
Gunderson JG, Morey LC, Stout RL, et al: Major depressive
disorder and borderline personality disorder revisited:
longitudinal associations. J Clin Psychiatry 65:1049
1056, 2004
Heatherton TF, Weinberger JL (eds): Can Personality Change?
Washington, DC, American Psychological Press, 1994
Johnson JG, Williams JBW, Goetz RR, et al: Stability and
change in personality disorder symptomatology: findings from a longitudinal study of HIV+ and HIV men.
J Abnorm Psychol 106:154158, 1997

113

Johnson JG, Cohen P, Smailes E, et al: Adolescent personality


disorders associated with violence and criminal behavior during adolescence and early adulthood. Am J Psychiatry 157:14061412, 2000a
Johnson JG, Cohen P, Kasen S, et al: Age-related change in
personality disorder trait levels between early adolescence and adulthood: a community-based longitudinal investigation. Acta Psychiatr Scand 102:265275,
2000b
Judd PH, McGlashan TH: A Developmental Model of Borderline Personality Disorder: Understanding Variations
in Course and Outcome. Washington, DC, American
Psychiatric Publishing, 2003
Katz MM, Secunda SK, Hirschfeld R, et al: NIMH Clinical
Research Branch Collaborative Program on Psychobiology of Depression. Arch Gen Psychiatry 36:765771,
1979
Keller MB, Shapiro RW, Lavori PW, et al: Recovery in major
depressive disorder: analysis with the lifetable and regression models. Arch Gen Psychiatry 39:905910, 1982
Keller MB, Lavori PW, Friedman B, et al: The Longitudinal
Interval Follow-Up Evaluation. Arch Gen Psychiatry
44:540548, 1987
Kendell PC, Clarkin JF: Introduction to special section: comorbidity and treatment implications. J Consult Clin
Psychol 60:833834, 1992
Klein DN, Shih JH: Depressive personality: associations
with DSM-III-R mood and personality disorders and
negative and positive affectivity, 30-month stability, and
prediction of course of axis I depressive disorders. J Abnorm Psychol 107:319327, 1998
Krueger RF: The structure of common mental disorders.
Arch Gen Psychiatry 56:921926, 1999
Krueger RF, Carlson SR: Personality disorders in children
and adolescents. Curr Psychiatry Rep 3:4651, 2001
Krueger RF, Tackett JL: Personality and psychopathology:
working toward the bigger picture. J Personal Disord
17:109128, 2003
Lenzenweger MF: Stability and change in personality disorder features: the longitudinal study of personality disorders. Arch Gen Psychiatry 56:10091015, 1999
Lenzenweger MF, Loranger AW, Korfine L, et al: Detecting
personality disorders in a nonclinical population: application of a 2-stage procedure for case identification.
Arch Gen Psychiatry 54:110133, 1997
Levy KN, Becker DF, Grilo CM, et al: Concurrent and predictive validity of the personality disorder diagnosis in adolescent inpatients. Am J Psychiatry 156:15221528,
1999
Lewinsohn PM, Rohde P, Seeley JR, et al: Axis II psychopathology as a function of Axis I disorder in childhood
and adolescents. J Am Acad Child Adolesc Psychiatry
36:17521759, 1997
Livesley WJ, Jackson DN, Schroeder ML: Factorial structure
of traits delineating personality disorders in clinical and
general population samples. J Abnorm Psychol 101:432
440, 1992

114

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Loranger AW. Personality Disorder Examination (PDE)


Manual. Yonkers, NY, DV Communications, 1988
Loranger AW, Lenzenweger MF, Gartner AF, et al: Trait-state
artifacts and the diagnosis of personality disorders.
Arch Gen Psychiatry 48:720728, 1991
Loranger AW, Sartorius N, Andreoli A, et al: The International Personality Disorder Examination (IPDE). Arch
Gen Psychiatry 51:215224, 1994
Lynam DR, Widiger TA: Using the five-factor model to represent the DSM-IV personality disorders: an expert consensus approach. J Abnorm Psychol 110:401412, 2001
Lyons MJ, Tyrer P, Gunderson J, et al: Special feature: heuristic models of comorbidity of Axis I and Axis II disorders. J Personal Disord 11:260269, 1997
Maddocks PD: A five year follow-up of untreated psychopaths. Br J Psychiatry 116:511515, 1970
Mattanah JJ, Becker DF, Levy KN, et al: Diagnostic stability
in adolescents followed up 2 years after hospitalization.
Am J Psychiatry 152:889894, 1995
McDavid JD, Pilkonis PA: The stability of personality disorder diagnoses. J Personal Disord 10:115, 1996
McGlashan TH: The Chestnut Lodge Follow-up Study, part
III: long-term outcome of borderline personalities. Arch
Gen Psychiatry 42:2030, 1986a
McGlashan TH: The Chestnut Lodge Follow-up Study, part
VI: schizotypal personality disorder. Arch Gen Psychiatry 43:329334, 1986b
McGlashan TH, Grilo CM, Skodol AE, et al: The Collaborative Longitudinal Personality Disorders Study: baseline
Axis I/II and II/II diagnostic co-occurrence. Acta Psychiatr Scand 102:256264, 2000
McGlashan TH, Grilo CM, Sanislow CA, et al: Two-year
prevalence and stability of individual DSM-IV criteria
for schizotypal, borderline, avoidant, and obsessivecompulsive disorders. Am J Psychiatry (in press)
Myers MG, Stewart DG, Brown SA: Progression from conduct disorder to antisocial personality disorder following treatment for adolescent substance abuse. Am J Psychiatry 155:479485, 1998
Nesselroade JR, Stigler SM, Baltes PB: Regression toward the
mean and the study of change. Psychol Bull 88:622637,
1980
Newman DL, Caspi A, Moffitt TE, et al: Antecedents of adult
interpersonal functioning: effects of individual differences in age three temperament. Dev Psychol 33:206
217, 1997
Oldham JM, Skodol AE, Kellman HD, et al: Diagnoses of
DSM-III-R personality disorders by two structured interviews: patterns of comorbidity. Am J Psychiatry 149:
213230, 1992
Ottosson H, Ekselius L, Grann M, et al: Cross-system concordance of personality disorder diagnoses of DSM-IV and
diagnostic criteria for research of ICD-10. J Personal
Disord 16:283292, 2002
Paris J: Personality disorders over time: precursors, course
and outcome. J Personal Disord 17:479488, 2003

Paris J, Zweig-Frank H: A twenty-seven year follow-up of


borderline patients. Compr Psychiatry 42:482487, 2001
Perry JC: Longitudinal studies of personality disorders.
J Personal Disord 7 (suppl):6385, 1993
Rettew DC, Zanarini MC, Yen S, et al: Childhood antecedents of avoidant personality disorder: a retrospective
study. J Am Acad Child Adolesc Psychiatry 42:1122
1130, 2003
Rey JM, Morris-Yates A, Singh M, et al: Continuities between
psychiatric disorders in adolescents and personality
disorders in young adults. Am J Psychiatry 152:895900,
1995
Robbins E, Gentry KA, Munoz RA, et al: A contrast of the
three more common illnesses with the ten less common
in a study and 18-month follow-up of 314 psychiatric
emergency room patients, III: findings at follow-up.
Arch Gen Psychiatry 34:285291, 1977
Roberts BW, DelVecchio WF: The rank-order consistency of
personality traits from childhood to old age: a quantitative review of longitudinal studies. Psychol Bull 126:3
25, 2000
Roberts BW, Caspi A, Moffitt TE: The kids are alright: growth
and stability in personality development from adolescence to adulthood. J Pers Soc Psychol 81:670683, 2001
Robins LN: Deviant Children Grown Up: A Sociological and
Psychiatric Study of Sociopathic Personality. Baltimore,
MD, Williams and Wilkins, 1966
Robins LN: Epidemiology: reflections on testing the validity
of psychiatric interviews. Arch Gen Psychiatry 42:918
924,1985
Ruegg R, Frances A: New research in personality disorders.
J Personal Disord 9:148, 1995
Sanislow CA, McGlashan TH: Treatment outcome of personality disorders. Can J Psychiatry 43:237250, 1998
Seivewright H, Tyrer P, Johnson T: Change in personality
status in neurotic disorders. Lancet 359:22532254, 2002
Shea MT, Yen S: Stability as a distinction between axis I and
axis II disorders. J Personal Disord 17:373386, 2003
Shea MT, Stout RL, Gunderson JG, et al: Short-term diagnostic stability of schizotypal, borderline, avoidant, and obsessive-compulsive personality disorders. Am J Psychiatry 159:20362041, 2002
Shea MT, Stout RL, Yen S, et al: Associations in the course of
personality disorders and Axis I disorders over time.
J Abnorm Psychol 113:449508, 2004
Siever LJ, Davis KL: A psychobiological perspective on the
personality disorders. Am J Psychiatry 148:16471658,
1991
Skodol AE: Classification, assessment, and differential diagnosis of personality disorders. J Pract Psychiatry Behav
Health 3:261274, 1997
Skodol AE, Pagano ME, Bender DS, et al: Stability of functional impairment in patients with schizotypal, borderline, avoidant, or obsessive-compulsive personality disorder over two years. Psychol Med (in press)

Course and Outcome of Personality Disorders

Solomon DA, Keller MB, Leon AC, et al: Recovery from major depression: a 10-year prospective follow-up across
multiple episodes. Arch Gen Psychiatry 54:10011006,
1997
Squires-Wheeler E, Skodol AE, Erlenmeyer-Kimling L: The
assessment of schizotypal features over two points in
time. Schizophrenia Res 6:7585, 1992
Stevenson J, Meares R, Comerford A: Diminished impulsivity in older patients with borderline personality disorder. Am J Psychiatry 160:165166, 2003
Stone MH: Long-term outcome in personality disorders. Br
J Psychiatry 162:299313, 1993
Trull TJ, Useda JD, Conforti K, et al: Borderline personality
disorder features in nonclinical young adults, 2: twoyear outcome. J Abnorm Psychol 106:307314, 1997
Trull TJ, Useda JD, Doan BT, et al: Two-year stability of borderline personality measures. J Personal Disord 12:187
197, 1998
Tyrer P: Personality disorder. Br J Psychiatry 179:8184, 2001
Tyrer P, Simonsen E: Personality disorder in psychiatric
practice. World Psychiatry 2:4144, 2003
Tyrer P, Strauss J, Cicchetti D: Temporal reliability of personality in psychiatric patients. Psychol Med 13:393398,
1983
Tyrer P, Gunderson J, Lyons M, et al: Special feature: extent
of comorbidity between mental state and personality
disorders. J Personal Disord 11:242259, 1997
Warner MB, Morey LC, Finch JF, et al: The longitudinal relationship of personality traits and disorders. J Abnorm
Psychol 113:217227, 2004

115

Westen D: Divergences between clinical and research methods for assessing personality disorders: implications for
research and the evolution of axis II. Am J Psychiatry
154:895903, 1997
Westen D, Shedler J: Revising and assessing axis II, Part II:
toward an empirically based and clinically useful classification of personality disorders. Am J Psychiatry 156:
273285, 1999
Widiger TA: Categorical versus dimensional classification:
implications from and for research. J Pers Disord 6:287
300, 1992
Widiger TA: Personality disorder and Axis I psychopathology: the problematic boundary of Axis I and Axis II.
J Personal Disord 17:90108, 2003
Wolff S, Townshend R, McGuire RJ, et al: Schizoid personality in childhood and adult life, II: adult adjustment and
the continuity with schizotypal personality disorder. Br
J Psychiatry 159:620629, 1991
World Health Organization. ICD-10: Classification of Mental
and Behavioral Disorders. Geneva, World Health Organization, 1992
Zanarini MC, Skodol AE, Bender D, et al: The Collaborative
Longitudinal Personality Disorders Study: reliability of
Axis I and II diagnoses. J Personal Disord 14:291299,
2000
Zanarini MC, Frankenburg FR, Hennen J, et al: The longitudinal course of borderline psychopathology: 6-year prospective follow-up of the phenomenology of borderline
personality disorder. Am J Psychiatry 160:274283, 2003
Zimmerman M: Diagnosing personality disorders: a review
of issues and research methods. Arch Gen Psychiatry
51:225245, 1994

This page intentionally left blank

Part III
Etiology

This page intentionally left blank

7
A Current Integrative
Perspective on Personality
Disorders
Joel Paris, M.D.

1991) is a general theory of psychopathology that is


both nonreductionistic and interactional. Every category of mental disorder is associated with some kind of
genetic vulnerability (Paris 1999). Yet genes are not the
direct causes of mental disorders; rather, they shape individual variability in temperament and traits. Some of
these temperamental variants constitute a vulnerability to psychopathology. By and large, however, traits
only become maladaptive under specific environmental conditions. In other words, diatheses become
apparent when uncovered by stressors. For example,
even in a condition such as schizophrenia, with its
well-established genetic risk, only half of identical
twins are concordant for the disorder (Meehl 1990).
The interactions between diatheses and stressors
are bidirectional. Genetic variability influences the way
individuals respond to their environment, while environmental factors determine whether genes are expressed. These relationships help explain why adverse
life events by themselves do not consistently lead to
pathological sequelae. Most children are resilient to all
but the most severe and consistent adversities (Rutter
and Maughan 1997). However, trauma, neglect, and

MENTAL DISORDERS AND THE STRESSDIATHESIS MODEL


By themselves, neither chemical imbalances, psychological adversities, nor troubled social environments
account for the development of psychopathology. A
multitude of interactions between biological, psychological, and social factors are involved in the etiology
of any mental disorder. This statement may be a truism, but in practice, we find it difficult to deal with
complexity. Although the real world is nonlinear and
multivariate, the human mind is structured for linear
thinking. Even researchers are not immune to oversimplifications.
One way to embrace complexity is to frame psychological phenomena in a systems perspective. General systems theory (Sameroff 1995) takes into account
the biological roots of behavior without reducing psychology to neurochemistry. Mental processes have
emergent properties that cannot be explained at other
levels of analysis.
The stress-diathesis model (Monroe and Simons
119

120

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

dysfunctional families probably have greater effects on


temperamentally vulnerable children (Paris 2000b).

TEMPERAMENT, TRAITS, AND


PERSONALITY DISORDERS
We can now apply these principles to the development of personality disorders. To conceptualize how
diatheses and stressors interact to shape personality
pathology, we need to consider the hierarchical and
nested relationship among personality disorders, personality traits, and temperament (Rutter 1987). Temperament reflects the genetic factors that account for a
large proportion of the variance in personality traits
(Plomin et al. 2001a). Personality traitsthat is, individual differences in behavior that remain stable over
time and contextare an amalgam of temperament
and experience. Personality disorders are dysfunctional outcomes that occur when these traits are amplified and used in rigid and maladaptive ways. By
themselves, trait differences are fully compatible with
normality, but trait profiles determine what type of
personality disorder can develop (Paris 2003).
Strong evidence supports the principle that there is
no definite boundary between personality traits and
personality disorders (Cloninger et al. 1993; Costa and
Widiger 2001; Livesley et al. 1993; Siever and Davis
1991). For this reason disorders are best understood as
pathological amplifications of traits (Millon and Davis
1995; Paris 2003). Whereas some personality disorders
(particularly those in the borderline category) show
symptoms that are rare in community populations,
they are still rooted in trait dimensions (Siever and
Davis 1991). Moreover, genetic, neuropsychological,
and biological markers are not consistently associated
with any of the categories of disorders described in
DSM-IV-TR (American Psychiatric Association 2000)
but are related to traits (Livesley 2003). Thus, traits are
closer to biological bedrock than disorders, which are
more colored by psychosocial influences. Similar principles can be broadly applied to all forms of psychopathology, including Axis I disorders, which are also
rooted in traits and temperament (Kendell and Jablensky 2003).
Epidemiological studies (Samuels et al. 2002; Torgersen et al. 2001; Weissman 1993) have estimated that
as much as 10% or more of the general population has
a personality disorder. However, if there is no absolute cutoff point between personality traits and disorders, one can question these findings. Although it is
probably true that one out of ten people has problem-

atic traits, everything depends on how much dysfunction is required to diagnose a disorder.
The overall criteria for diagnosis of a personality
disorder in DSM-IV-TR require an enduring pattern of
inner experience and behavior that deviates markedly
from the expectations of the individuals culture, affecting cognition, affect, interpersonal functioning, and
impulse control. The pattern must be inflexible and
pervasive across a broad range of personal and social
situations, leading to clinically significant distress or
impairment in social, occupational, or other important
areas of functioning. Finally, the pattern must be stable
and of long duration and must have an onset that can
be traced back to adolescence or early adulthood.
Each of these criteria requires an informed clinical
judgment. For this reason, personality disorder diagnoses may not be reliable unless pathology is severe.
Moreover, only three categories of personality disorder have a large empirical literature (schizotypal, antisocial, and borderline). Given the less well-defined
characteristics of the other categories, it is not surprising that many patients meet overall criteria in DSMIV-TR for a personality disorder but do not fall into
any specific category and can only be classified as personality disorder not otherwise specified. About a
third of all cases in practice fall into this group (Loranger et al. 1994).
Ever since the publication of DSM-III (American
Psychiatric Association 1980), psychiatry has classified mental disorders on the basis of phenomenology.
This decision was the right one for its time. In the absence of solid data on etiology, it is better to categorize
what clinicians can observe. However, an etiologically
based classification must be our ultimate goal. In this
respect, all the categories of disorder on Axis II of
DSM-IV-TR can only be considered provisional. Future classifications of personality disorders may be
based on the underlying neurobiological mechanisms
that shape traits (Paris 2000a). This approach, which
involves classifying disease on the basis of pathogenesis or etiology, is becoming standard in all areas of
medicine. It would combine data at many levels of
analysis: molecular genetics, behavior genetics, neurochemistry, neurophysiology, cognitive science, and
developmental psychopathology. Classifying personality disorders in this way need not exclude the crucial
influence of environmental factors, which ultimately
affect brain circuitry and brain chemistry.
Factor and cluster analytic studies of personality
traits and disorders suggest an underlying structure
that is obscured by current diagnoses (Livesley 2003).
Several suggestions have been made about the nature

A Current Integrative Perspective on Personality Disorders

of this structure. Costa and Widiger (2001) have proposed that disorders can be accounted for by the fivefactor model (FFM) of personality. Livesley and Jang
(2000) have developed a somewhat similar model,
with superfactors that parallel four of the five factors
in the FFM. Cloninger et al. (1993) have developed a
seven-factor model that also describes similar trait dimensions.
Applying any of these systems would lead to a
very different classification of personality disorders. A
dimensional system would help to deal with widespread comorbidity of Axis II disorders. Many (albeit
not all) of these overlaps occur within the Axis II clusters (Pfohl et al. 1986), supporting the concept that
trait dimensions underlie categories. The problem is
that current categories of personality disorders are not
well-defined phenotypes (Jang et al. 2001); this is
probably why diagnoses tend to overlap.
Krueger (1999) found that almost all DSM-defined
disorders can be accounted for by factors that reflect
internalizing and externalizing symptoms; these are the
same superfactors that emerge from studies of psychopathology in children (Achenbach and McConaughy 1997). These broad factors also correspond to
the personality trait dimensions measured by the FFM
in adult community populations (Costa and Widiger
2001): internalizing dimensions are associated with
high levels of introversion and neuroticism, whereas
externalizing dimensions are associated with high extraversion and low conscientiousness. However, this
distinction fails to take into account another crucial
trait for psychopathology: the cognitive dimension associated with vulnerability to psychotic disorders.
Building dimensional models from psychopathology
rather than from normality allows us to include phenomena rarely seen in community populations. For
example, in a model linking traits and overt disorders,
Siever and Davis (1991) conceptualized all categories
(on both Axis I and Axis II) within four trait dimensions: cognitive, depressive, impulsive, and anxious.
The Axis II clusters described in DSM-IV-TR are
rough-and-ready clinically derived concepts and need
to be redefined to improve their boundaries. At this
point, the Axis II clusters are only approximations of
spectra that include a number of overlapping disorders. Nonetheless, the clusters show some interesting
parallels with dimensional approaches to personality
disorders.
The categories in Cluster A (schizoid, paranoid,
and schizotypal) are related to the schizophrenia spectrum (Paris 2003; Siever and Davis 1991). Similarly, categories in Cluster B are associated with trait impulsiv-

121

ity and/or affective dysregulation (Siever and Davis


1991; Zanarini 1993). This is most clearly apparent for
antisocial personality disorder (ASPD) and borderline
personality disorder (BPD), but histrionic and narcissistic personality also show these features, albeit in attenuated form (Looper and Paris 2000). The situation
for Cluster C is more complex: although avoidant and
dependent personality disorders are clearly associated
with trait anxiety (Kagan 1994; Paris 1997), the compulsive category may reflect a separate compulsivity
trait dimension (Livesley and Jang 2000).

GENES, ENVIRONMENT, AND


PERSONALITY TRAITS
It is difficult to separate the influence of genes from
that of environment on personality. It has been consistently shown that personality traits are heritable, with
genetic factors accounting for nearly half of the variance (Plomin et al. 2001a). However, single genes are
not associated with single temperamental characteristics. Rather, the heritable component of personality
emerges from complex and interactive polygenetic
mechanisms associated with variations in multiple allelesthat is, quantitative trait loci (Rutter and Plomin 1997). Thus, attempts to find genetic associations
between genes and traits have been disappointing,
and research needs to take environmental effects on
gene expression into account (Rutter et al. 2001).
The existence of a genetic component in personality suggests that traits may be linked to biological
markers. These relationships have remained rather
obscure. The most robust finding in the literature is a
relationship between low levels of central serotonin
activity and impulsivity (Mann 1998). Again, the problem lies in the lack of precisely defined phenotypes.
Livesley (2003) suggested that genes and biological
markers are more likely to correlate with narrowly defined traits (which may be affected by fewer alleles)
than with broader traits.
In behavioral genetic research, half of the variance
in personality traits derives from the environment, but
this portion is almost entirely unshared (Plomin et
al. 2001a, 2001b). These findings show that environmental influences on traits do not necessarily derive
from being raised in the same family. This finding has
been the subject of great controversy, because it contradicts many classical ideas in developmental and
clinical psychology that focus on parenting as a primary factor shaping personality development (Harris
1998; Paris 2000b).

122

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

There are several possible reasons why unshared


(but not shared) environmental factors are important
in personality (Plomin et al. 2001b). The first is that a
childs temperament affects the response of other people in the childs environment. In a large-scale study of
adolescents (Reiss et al. 2000) that used a combination
of twin and family methods, multivariate analyses
showed that the temperament of the child was the underlying factor driving differential parenting and differential behavioral outcomes.
A second explanation is that even when the family
provides a similar environment for siblings, each child
may perceive experiences differently and respond to
them with different behavioral patterns. Again, temperamental differences can make environmental influences unshared.
A third explanation is that some environmental
factors affecting personality are extrafamilial. Every
child has shaping experiences with peers, with teachers, or with community leaders (Rutter and Maughan
1997). Harris (1998) proposed that peer groups might
be even more crucial than parents for personality development.
A final possibility is that personality might be affected by intrauterine factors, a biological environment
that is not shared. However, there is little research supporting this hypothesis.
Whatever the ultimate explanation, almost all the
empirical literature claiming to establish links between
childhood experiences and personality has to be questioned on the grounds that genetic factors may be latent
variables accounting for some of these relationships
(Harris 1998). This difficulty also affects the validity of
research measures. Plomin and Bergeman (1991) reported that behavioral genetic studies of standard measures of life experience, past and present, have a heritable component that correlates with personality trait
differences.

Genetic Factors
If personality disorders are pathologically amplified
traits, it would be logical for them to show heritability
levels similar to personality dimensions. This expectation has been confirmed empirically. Torgersen et al.
(2000) located a large sample of twins in Norway in
which one proband met criteria for at least one categorical Axis II diagnosis (except for the antisocial category). All personality disorders had heritabilities resembling those observed for traits (i.e., close to half the
variance). Although the findings cannot be considered
quantitatively precise (in view of small sample size

and local variations in base rates), they were consistent across disorders. Moreover, even in personality
disorders that have not traditionally been considered
heritable (such as the borderline and narcissistic categories), genetic factors accounted for more than half of
the variance. Although there were no patients with
ASPD in the Norwegian cohort, other lines of research
(Cloninger et al. 1982) have pointed to heritable influences on that disorder.
Genetic factors influencing both traits and disorders have been supported by family history studies examining spectra of disorders across Axis I and Axis II.
Thus, first-degree relatives of patients with disorders
in Cluster A show pathology in the schizophrenic spectrum (Siever and Davis 1991); those of patients in Cluster B tend to have other impulsive disorders (Zanarini
1993) in some cases and other affective disorders in
other cases (Siever and Davis 1991); relatives of patients in Cluster C often have anxiety disorders (Paris
1997). These findings support the relationship of trait
dimensions to disorders. Moreover, if these same traits
underlie all forms of psychopathology, it should not be
surprising that patients with Axis II disorders can have
wide-ranging Axis I comorbidity (McGlashan et al.
2000; Zanarini et al. 2001).
The influence of genetic factors on personality disorders supports a continued search for biological markers associated with Axis II disorders and their underlying traits. In Cluster A disorders, biological markers,
such as abnormal eye tracking, are found that are also
common in schizophrenia (Siever and Davis 1991). In
Cluster B disorders, we see the same relationship between central serotonin activity and impulsive aggression that has been studied on the trait level (Coccaro et
al. 1989). The most consistent results in neurophysiological and neuropsychological research on Cluster B
disorders are also related to impulsive traits. Thus,
functional abnormalities in prefrontal cortex are associated with impulsive aggression, as shown by decreases
in the mass of frontal gray matter in subjects with ASPD
(Raine et al. 2000). Patients with ASPD and BPD have
deficits in executive function as measured by the Wisconsin Card Sorting Test (OLeary 2000). Although we
know much less about Cluster C disorders, the physiological correlates of trait anxiety have been measured in
longitudinal designs (Kagan 1994).

Psychological Factors
A large body of evidence supports the concept that
childhood adversities are risk factors for personality
disorders (Paris 2003). The problem is that there is in-

A Current Integrative Perspective on Personality Disorders

sufficient evidence to establish a direct causal relationship. For example, research on patients with BPD has
documented that histories of sexual abuse, physical
abuse, and gross neglect are common (Paris 1994; Zanarini 2000). One current theory of BPD is that children who develop this disorder have abnormal patterns of attachment that emerge from exposure to
adversity (Fonagy et al. 1995). However, longitudinal
studies are needed to determine the origins of these
patterns as well as their impact on development.
It has been consistently shown that the impact of
childhood adversities is different in clinical and community samples. Community surveys of the effects of
childhood sexual abuse (Browne and Finkelhor 1986;
Rind and Tromofovitch 1997), as well as of physical
abuse (Malinovsky-Rummell and Hansen 1993), have
found that only a minority of children exposed to abuse
and trauma suffer measurable sequelae. One explanation could be that adverse life experiences lead to psychopathology only in the presence of specific trait profiles associated with temperamental vulnerability.
Finally, single traumatic events are rarely, by themselves, associated with pathological sequelae; instead,
continuously adverse circumstances have cumulative
effects associated with the development of symptomatology (Rutter 1989). For this reason, one cannot understand the impact of childhood adversities without
placing events within a longitudinal and developmental context.
Another problem with the existing research literature is that most studies have examined childhood risk
factors using retrospective methodologies. Reports of
life experiences occurring many years in the past tend
to be colored by recall biasthat is, the tendency for
individuals with symptoms in the present to remember more adversities in the past (Robins et al. 1985;
Schacter 1996). To address this problem we need longitudinal data.
A good example of the kind of study we require is
the follow-back study by Robins (1966) of children
with conduct disorder. Here it was observed that the
strongest predictor of adult ASPD among children
with conduct disorder was parental psychopathy
(usually in the father), an association later supported
by Farrington (1998). Several studies have demonstrated that first-degree relatives of patients with BPD
have increased levels of impulsive spectrum disorders
(Links et al. 1988; Zanarini 1993).
Yet even here, causality is unclear. Because the mechanism behind these relationships could involve inheritance, modeling, or pathological parenting, one needs to
separate the effects of personality traits common be-

123

tween parents and children from the effects of family


dysfunction. For this reason, research methods are
needed in which temperament is controlled for using
behavior-genetic designs. An ongoing study (Dionne et
al. 2003) has been prospectively following large cohorts
of monozygotic and dizygotic twins from infancy, but
these cohorts have only reached middle childhood.
Community studies avoid the problems in studying
clinical populations because clinical samples are already biased toward psychopathology. One large-scale
prospective longitudinal project, the Albany-Saratoga
study, has been following a cohort of children from
middle childhood to early adulthood and examining
the predictors of pathological sequelae. Johnson et al.
(1999) reported that early adversities, including neglect,
physical abuse, and sexual abuse, were associated with
a higher number of personality disorder symptoms.
This study is important and unique, but the researchers
had to use a continuous variable to measure outcome
because there were not enough subjects with a diagnosable personality disorder. Also, the research design
lacked data on temperamental factors in early childhood that might have preceded environmental adversities and affected their impact.
With all these caveats, it is impossible to escape the
conclusion that adversity in childhood, as well as in
later life, is a crucial factor affecting the development
of personality disorders. The impact of negative life
events tends to amplify temperamental vulnerability
(Paris 1996) but can often be modulated by resilience
factors (Rutter 1989).

Social Factors
The role of social factors on personality disorders has
not been widely researched. Yet, like other forms of
psychopathology, personality disorders develop in a
sociocultural context. There are two ways to test this
relationship. First, one could look for cross-cultural
differences in personality disorders (Paris 1996). Second, one could determine whether personality disorders vary in prevalence over time (Paris 2004).
Mental disorders can present with different symptoms in different cultures, and some categories of illness are seen only in specific social settings (Murphy
1982). Personality disorders are socially sensitive
(Paris 2004) because their symptoms reflect behaviors
and feelings that could be shaped and molded by culture. Moreover, if traits themselves show sociocultural
variation, personality disorders might present with
different symptoms in different social contexts, and
some categories might even be culture bound.

124

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

The broader dimensions of personality have been


shown to be similar in different societies (McCrae and
Costa 1999), but this may not be the case for personality
disorders. Whereas Loranger et al. (1994) showed that
the categories in DSM-IV-TR and ICD-10 (World Health
Organization 1992) can be identified in clinical settings
around the world, there are no epidemiological data
concerning possible differences in prevalence between
cultures and societies. This lacuna may be partially
filled by the upcoming International Comorbidity
Study, which will be a replication of the National Comorbidity Survey (Kessler et al. 1994) in several countries. The largest community surveys, such as the Epidemiological Catchment Area Study (Robins and
Regier 1991) as well as the National Comorbidity Survey, have examined only ASPD (which has behavioral
symptoms that are readily measured). The International Comorbidity Study will make use of a reliable instrument, the International Personality Disorder Examination (Loranger et al. 1994), that will also determine
the prevalence of BPD.
At this point, cross-cultural studies support the role
of social factors in ASPD. This category is relatively rare
in traditional societies such as Taiwan (Hwu et al. 1989)
and Japan (Sato and Takeichi 1993), but prevalence
reaches North American and European levels in Korea
(Lee et al. 1987). The East Asian cultures that have a low
prevalence of ASPD have cultural and family structures
that are protective against antisocial behavior. Thus,
families are a mirror image of the risk factors for the disorder described by Robins (1966): fathers are strong and
authoritative, expectations of children are high, and
family loyalty is prized. In addition, communities outside the family have high social cohesion, further containing those with impulsive temperament. One might
also hypothesize that less well-structured family and
social structures are among the factors that make ASPD
more common in Western societies.
The strongest evidence thus far for sociocultural
factors in personality disorders comes from cohort effects (changes in prevalence over short periods of
time). ASPD (as well as other impulsive spectrum disorders such as substance abuse) has become more
common in adolescents and young adults, both in
North America and Europe, since World War II (Rutter
and Smith 1995).
There may also be cohort effects on the prevalence
of BPD (Millon 2000; Chapter 14, Sociocultural Factors). Expanding on this thesis, Paris (1996) pointed to
several lines of supportive evidence: recent increases in
the prevalence of parasuicide and completed suicide
(Bland et al. 1998) and the observation that a third of

youths who commit suicide can be diagnosed with BPD


(Lesage et al. 1994).
One likely mechanism for an increase of prevalence
in BPD may derive from the breakdown of traditional
structures guiding the development of adolescents and
young adults (Millon 2000). Traditional societies have
long been defined in the sociological literature (e.g.,
Lerner 1958) as having high social cohesion, fixed social
roles, and high intergenerational continuity; these characteristics stand in contrast to modern societies, which
have lower social cohesion, fluid social roles, and lower
continuity between generations. Although traditional
societies could carry a different set of risks for psychopathology, the problem of identity is often associated
with personality disorders and may be exacerbated by
the conditions of modernity in which individuals must
develop their own social roles (Paris 1996).
Impulsive disorders (substance abuse, eating disorders, ASPD, BPD) may be particularly responsive to
social context because they are contained by structure
and limits and amplified by the absence of these.
However, these conditions would only be expected to
develop in individuals who also have the biological
and psychological risk factors for impulsive disorders.
Linehan (1993) suggested that patients with BPD act
impulsively as way of dealing with emotional dysregulation and that decreases in social support in modern
society amplify these traits by interfering with a buffering mechanism.
The relationship of social factors is less clear for
other personality disorders. In narcissistic personality
disorder, one might hypothesize that underlying traits
may no longer be channeled into fruitful ambition due
to breakdowns in family and social structures (Kohut
1977; Paris 2003). Similarly, avoidant personality disorder might be understood as reflecting the outcome
of social anxiety in modern society. Kagan (1994) has
studied behavioral inhibition in infants, a temperamental syndrome that increases the risk for anxiety
disorders later in life. In a traditional society, anxious
traits would be buffered by family and community
structures, whereas in modern society, the same traits
are more likely to become disabling and lead to disorders (Paris 1997).

IMPLICATIONS OF A STRESS-DIATHESIS
MODEL OF PERSONALITY DISORDERS
The biological, psychological, and social risk factors for
personality disorders can be integrated within a stress-

A Current Integrative Perspective on Personality Disorders

diathesis model. Both genetic-temperamental and psychosocial factors would be necessary conditions for the
development of personality disorders, but neither
would be sufficient. A combination of risksthat is, a
two-hit or multiple-hit mechanismis required.
The effects of psychosocial adversity will be greatest in
individuals who are temperamentally predisposed to
psychopathology. The cumulative effects of multiple
risk factors, rather than single adversities, will determine whether psychopathology develops. Finally, the
specific disorder that emerges depends on temperamental profiles specific to the individual.
Geneenvironment interactions would further mediate the pathogenesis of personality disorders. Abnormal temperament is associated with a greater sensitivity to environmental risk factors, and children with
problematic temperaments are more likely to experience adversities (Rutter and Maughan 1997). Vulnerable children also elicit responses from others that tend
to amplify their most problematic characteristics, creating a positive feedback loop. These adverse experiences further amplify traits, increasing the risk for further adversities.
An integrative model also helps to account for the
course of personality disorders over time (Paris 2003).
Early onset of pathology probably tends to reflect abnormal temperament. ASPD is a good example: even
as early as age 3, behavioral disturbances predict its development in adulthood (Caspi et al. 1996; Kim-Cohen
et al. 2003; Zoccolillo et al. 1992). However, the development of conduct symptoms in childhood is clearly
related to family pathology (Patterson and Yoerger
1997; Robins 1966). Similarly, children with unusual
shyness and reactivity may be at higher risk for anxious cluster personality disorders (Paris 1998), but
these traits may be amplified by family experience
(Head et al. 1991).
By adolescence, when personality trait patterns become stable (Costa and McCrae 2001), one can diagnose typical cases of personality disorder (Kernberg et
al. 2000), although specific categories tend to shift over
time (Bernstein et al. 1993). In adult life, most personality disorders have a chronic course (Seivewright et
al. 2002), possibly due to continuing interactions between temperament and experience. However, Cluster B disorders are the exception because they tend to
burn out by middle age (Paris 2003), possibly reflecting the evolution of traits, with impulsivity leveling out over time.
A stress-diathesis model of personality disorders
also has important implications for treatment. It suggests that neither a purely biological or a purely psy-

125

chosocial perspective is a useful guide to effective


treatment of personality disorders. A strictly biological perspective tends to support a strongly pharmacological approach to these patients. Yet clinical trials do
not show specific efficacy for existing drugs, even if
some produce a certain degree of symptomatic relief
(Paris 2003; Soloff 2000). The limitations of pharmacotherapy are shown by evidence that patients with BPD
may be given as many as four or five drugs (Zanarini
et al. 2001) despite the fact that polypharmacy rarely
yields dramatic results.
A psychosocial perspective on personality disorders has generally supported psychotherapy as a primary form of treatment. However, maladaptive traits,
established in childhood and reinforced during adult
life, are often difficult to change. The efficacy of longterm therapies has not been established in clinical trials. Patients with personality disorders are generally
less responsive to standard forms of psychotherapy
than are patients with Axis I disorders without any
Axis II comorbidity (Shea et al. 1990). Psychodynamic
therapy has been tested in selected patient populations, most particularly those with BPD (Bateman and
Fonagy 1999; Stevenson and Meares 1992), but it is not
known whether these results are specific to the techniques used and whether they are generalizable to ordinary practice. Cognitive approaches to personality
disorders (Beck and Freeman 1990) have generated investigation, and dialectical behavior therapy for BPD
(Linehan 1993) has been consistently shown to reduce
impulsive behavior in BPD but has not been examined
for long-term efficacy.
Personality disorders will probably remain difficult to treat until we understand their etiology and
pathogenesis. As we obtain more knowledge concerning the diatheses and stressors driving both traits and
disorders, we will be in a better position to develop
more specific and more useful forms of treatment for
these patientsmore targeted biological interventions
and more targeted forms of psychotherapy.

REFERENCES
Achenbach TM, McConaughy SH: Empirically Based Assessment of Child and Adolescent Psychopathology:
Practical Applications, 2nd Edition. Thousand Oaks,
CA, Sage, 1997
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition. Washington,
DC, American Psychiatric Association, 1980

126

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

American Psychiatric Association: Diagnostic and Statistical


Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association,
2000
Bateman A, Fonagy P: Effectiveness of partial hospitalization
in the treatment of borderline personality disorder: a randomized controlled trial. Am J Psychiatry 156:15631569,
1999
Beck AT, Freeman A: Cognitive Therapy of Personality Disorders. New York, Guilford, 1990
Bernstein DP, Cohen P, Skodol A, et al: Prevalence and stability of the DSM-III personality disorders in a communitybased survey of adolescents. Am J Psychiatry 150:1237
1243, 1993
Bland RC, Dyck RJ, Newman SC, et al: Attempted suicide in
Edmonton, in Suicide in Canada. Edited by Leenaars
AA, Wenckstern S, Sakinofsky I, et al. Toronto, Canada,
University of Toronto Press, 1998, pp 136150
Browne A, Finkelhor D: Impact of child sexual abuse: a review of the literature. Psychol Bull 99:6677, 1986
Caspi A, Moffitt TE, Newman DL, et al: Behavioral observations at age three predict adult psychiatric disorders:
longitudinal evidence from a birth cohort. Arch Gen
Psychiatry 53:10331039, 1996
Cloninger CR, Sigvardsson S, Bohman M, et al: Predisposition to petty criminality in Swedish adoptees, II: crossfostering analysis of geneenvironment interactions.
Arch Gen Psychiatry 39:12421247, 1982
Cloninger CR, Svrakic DM, Pryzbeck TR: A psychobiological model of temperament and character. Arch Gen Psychiatry 50:975990, 1993
Coccaro EF, Siever LJ, Klar HM, et al: Serotonergic studies in
patients with affective and personality disorders. Arch
Gen Psychiatry 46:587599, 1989
Costa PT, Widiger TA (eds): Personality Disorders and the
Five-Factor Model of Personality, 2nd Edition. Washington, DC, American Psychological Association, 2001
Dionne G, Tremblay R, Boivin M, et al: Physical aggression
and expressive vocabulary in 19-month-old twins. Dev
Psychol 39:261273, 2003
Farrington DP: Youth crime and antisocial behavior, in The
Social Child. Edited by Campbell A, Muncer S. Hove,
England, Psychology Press, 1998, pp 353392
Fonagy P, Steele M, Steele H, et al: Attachment, the reflective
self, and borderline states, in Attachment Theory: Social, Developmental and Clinical Perspectives. Edited
by Goldberg S, Muir R, Kerr J. Hillsdale, NJ, Analytic
Press, 1995, pp 233278
Harris JR: The Nurture Assumption. New York, Free Press,
1998
Head SB, Baker JD, Williamson DA: Family environment
characteristics and dependent personality disorder.
J Personal Disord 5:256263, 1991
Hwu HG, Yeh EK, Change LY: Prevalence of psychiatric disorders in Taiwan defined by the Chinese Diagnostic Interview Schedule. Acta Psychiatr Scand 79:136147, 1989

Jang K, Vernon PA, Livesley WJ: Behavioural Genetic Perspectives on Personality Function. Can J Psychiatry
46:234244, 2001
Johnson JJ, Cohen P, Brown J, et al: Childhood maltreatment
increases risk for personality disorders during early
adulthood. Arch Gen Psychiatry 56:600606, 1999
Kagan J: Galens Prophecy. New York, Basic Books, 1994
Kendell R, Jablensky A: Distinguishing between the validity
and utility of psychiatric diagnoses. Am J Psychiatry
160:412, 2003
Kernberg PF, Weiner AS, Bardenstein KK: Personality Disorders in Children and Adolescents. New York, Basic
Books, 2000
Kessler RC, McGonagle KA, Nelson CB, et al: Lifetime and
12-month prevalence of DSM-III-R psychiatric disorders in the United States. Arch Gen Psychiatry 51:814,
1994
Kim-Cohen J, Caspi A, Moffitt TE, et al: Prior juvenile diagnoses in adults with mental disorder: developmental
follow-back of a prospective-longitudinal cohort. Arch
Gen Psychiatry 60:709717, 2003
Kohut H: The Restoration of the Self. New York, International Universities Press, 1977
Krueger RF: The structure of common mental disorders.
Arch Gen Psychiatry 56:921926, 1999
Lee KC, Kovac YS, Rhee H: The national epidemiological
study of mental disorders in Korea. J Korean Med Sci
2:1934, 1987
Lerner D: The Passing of Traditional Society. New York, Free
Press, 1958
Lesage AD, Boyer R, Grunberg F, et al: Suicide and mental
disorders: a case control study of young men. Am J Psychiatry 151:10631068, 1994
Linehan MM: Cognitive Behavioral Therapy of Borderline
Personality Disorder. New York, Guilford, 1993
Links PS, Steiner B, Huxley G: The occurrence of borderline
personality disorder in the families of borderline patients. J Personal Disord 2:1420, 1988
Livesley WJ: Personality Disorders: A Practical Approach.
New York, Guilford, 2003
Livesley WJ, Jang KL: Toward an empirically based classification of personality disorder. J Personal Disord 14:137
151, 2000
Livesley WJ, Jang KL, Jackson DN, et al: Genetic and environmental contributions to dimensions of personality
disorders. Am J Psychiatry 150:18261831, 1993
Looper K, Paris J: What are the dimensions underlying Cluster B personality disorders? Compr Psychiatry 41:432
437, 2000
Loranger AW, Sartori N, Andreoli A, et al: The International
Personality Disorder Examination. Arch Gen Psychiatry 51:215224, 1994
Malinovsky-Rummell R, Hansen DJ: Long-term consequences of physical abuse. Psychol Bull 114:6879, 1993
Mann JJ: The neurobiology of suicide. Nat Med 4:425430,
1998

A Current Integrative Perspective on Personality Disorders

McCrae RR, Costa PT: A five-factor theory of personality, in


Handbook of Personality: Theory and Research, 2nd
Edition. Edited by Pervin LA, John OP. New York, Guilford, 1999, pp 139153
McGlashan TH, Grilo CM, Skodol AE, et al: The Collaborative Longitudinal Personality Disorders Study: baseline
Axis I/II and II/II diagnostic co-occurrence. Acta Psychiatr Scand 102:256264, 2000
Meehl PE: Toward an integrated theory of schizotaxa, schizotypy, and schizophrenia. J Personal Disord 4:199, 1990
Millon T: Sociocultural conceptions of the borderline personality. Psychiatr Clin North Am 23:123136, 2000
Millon T, Davis R: Personality Disorders: DSM-IV and Beyond. New York, Wiley, 1995
Monroe SM, Simons AD: Diathesis-stress theories in the context of life stress research. Psychol Bull 110:406425,
1991
Murphy HBM: Comparative Psychiatry. New York, Springer,
1982
OLeary KM: Neuropsychological testing results. Psychiatr
Clin North Am 423:160, 2000
Paris J: Borderline Personality Disorder: A Multidimensional Approach. Washington, DC, American Psychiatric Press, 1994
Paris J: Social Factors in the Personality Disorders. Cambridge, England, Cambridge University Press, 1996
Paris J: Childhood trauma as an etiological factor in the personality disorders. J Personal Disord 11:3449, 1997
Paris J: Anxious traits, anxious attachment, and anxious cluster personality disorders. Harv Rev Psychiatry 6:142
148, 1998
Paris J: Nature and Nurture in Psychiatry. Washington, DC,
American Psychiatric Press, 1999
Paris J: The classification of personality disorders should be
rooted in biology. J Personal Disord 14:127136, 2000a
Paris J: Myths of Childhood. Philadelphia, PA, Brunner/
Mazel, 2000b
Paris J: Personality Disorders Over Time. Washington, DC,
American Psychiatric Publishing, 2003
Paris J: Sociocultural factors in the management of personality disorders, in Treatment of Personality Disorders. Edited by Magnavita T. New York, Wiley, 2004, pp 135147
Patterson GR, Yoerger K: A developmental model for lateonset delinquency. Nebr Symp Motiv 44:119177, 1997
Pfohl B, Coryell W, Zimmerman M, et al: DSM-III personality disorders: diagnostic overlap and internal consistency of individual DSM-III criteria. Compr Psychiatry
27:2134, 1986
Plomin R, Bergeman CS: The nature of nurture: genetic influence on environmental measures. Behav Brain Sci
14:373427, 1991
Plomin R, DeFries JC, McClearn GE, et al: Behavioral Genetics, 4th Edition. New York, Freeman, 2001a
Plomin R, Asbury K, Dunn J: Why are children in the same
family so different? Nonshared environment a decade
later. Can J Psychiatry 46:225233, 2001b

127

Raine A, Lencz T, Bilhul S: Reduced prefrontal gray matter


and reduced autonomic activity in antisocial personality disorder. Arch Gen Psychiatry 37:119127, 2000
Reiss D, Hetherington EM, Plomin R: The Relationship
Code. Cambridge, MA, Harvard University Press, 2000
Rind B, Tromofovitch P: A meta-analytic review of findings
from national samples on psychological correlates of
child sexual abuse. J Sex Res 34:237255, 1997
Robins LN: Deviant Children Grown Up. Baltimore, MD,
Williams and Wilkins, 1966
Robins LN, Regier DA (eds): Psychiatric Disorders in America. New York, Free Press, 1991
Robins LN, Schoenberg SP, Holmes SJ: Early home environment and retrospective recall: a test for concordance between siblings with and without psychiatric disorders.
Am J Orthopsychiatry 55:2741, 1985
Rutter M: Temperament, personality, and personality development. Br J Psychiatry 150:443448, 1987
Rutter M: Pathways from childhood to adult life. J Child
Psychol Psychiatry 30:2351, 1989
Rutter M, Maughan B: Psychosocial adversities in psychopathology. J Personal Disord 11:1933, 1997
Rutter M, Plomin R: Opportunities for psychiatry from genetic findings. Br J Psychiatry 171:209219, 1997
Rutter M, Smith DJ: Psychosocial Problems in Young People.
Cambridge UK, Cambridge University Press, 1995
Rutter M, Pickles A, Murray R, et al: Testing hypotheses on
specific environmental causal effects on behavior. Psychol Bull 127:291324, 2001
Sameroff AJ: General systems theories and developmental
psychopathology, in Developmental Psychopathology:
Theory and Methods. Edited by Ciccheti D, Cohen DJ.
New York, Wiley, 1995, pp 659699
Samuels J, Eaton WW, Bienvenu J, et al: Prevalence and correlates of personality disorders in a community sample.
Br J Psychiatry 180:536542, 2002
Sato T, Takeichi M: Lifetime prevalence of specific psychiatric disorders in a general medicine clinic. Gen Hosp
Psychiatry 15:224233, 1993
Schacter DL: Searching for Memory: The Brain, the Mind,
and the Past. New York, Basic Books, 1996
Seivewright H, Tyrer P, Johnson T: Change in personality
status in neurotic disorders. Lancet 359:22532254, 2002
Shea MT, Pilkonis PA, Beckham E: Personality disorders and
treatment outcome in the NIMH Treatment of Depression Collaborative Research Program. Am J Psychiatry
147:711718, 1990
Siever LJ, Davis KL: A psychobiological perspective on the
personality disorders. Am J Psychiatry 148:16471658,
1991
Soloff P: Psychopharmacological treatment of borderline
personality disorder. Psychiatr Clin North Am 23:169
192, 2000
Stevenson J, Meares R: An outcome study of psychotherapy
for patients with borderline personality disorder. Am
J Psychiatry 149:358362, 1992

128

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Torgersen S, Lygren S, Oien PA, et al: A twin study of personality disorders. Compr Psychiatry 41:416425, 2000
Torgersen S, Kringlen E, Cramer V: The prevalence of personality disorders in a community sample. Arch Gen
Psychiatry 58:590596, 2001
Weissman MM: The epidemiology of personality disorders:
a 1990 update. J Personal Disord 3 (suppl 7):4462, 1993
World Health Organization: International Classification of
Diseases, 10th Edition. Geneva, Switzerland, World
Health Organization, 1992
Zanarini MC: Borderline personality as an impulse spectrum disorder, in Borderline Personality Disorder: Etiology and Treatment. Edited by Paris J. Washington, DC,
American Psychiatric Press, 1993, pp 6786

Zanarini MC: Childhood experiences associated with the development of borderline personality disorder. Psychiatr
Clin North Am 23:89101, 2000
Zanarini MC, Frankenburg FR, Khera GS, et al: Treatment
histories of borderline inpatients. Compr Psychiatry
42:144150, 2001
Zoccolillo M, Pickles A, Quinton D, et al: The outcome of
childhood conduct disorder: implications for defining
adult personality disorder and conduct disorder. Psychol Med 22:971986, 1992

8
Epidemiology
Svenn Torgersen, Ph.D.

From clinical work we get an impression of which personality disorders are more common and which are
rarer. However, people with some types of personality disorders may be more likely to seek treatment and
obtain treatment compared with people with other
types of personality disorders. Consequently, if we are
interested in how prevalent different personality disorders are in the general population, we have to study
representative samples of the general population. Epidemiological research does just that.
Clinical work also gives us ideas about relationships between socioeconomic and sociodemographic
factors and personality disorders. However, in a clinical setting we only meet those from an unfavorable environment who have developed a personality disorder.
We do not meet those from an unfavorable environment who have not developed a disorder. Furthermore,
the combination of a specific personality disorder and
specific sociodemographic features may increase the
likelihood of a particular person to seek treatment.
These complexities mean that only population (epidemiological) studies can demonstrate the true relationship between personality disorders and socioeconomic and sociodemographic variables, or any other
variables such as traumata, disastrous events, upbringing, or partner relationships.

PREVALENCE
We know much about the prevalence of Axis I disorders in the general population (Kringlen et al. 2001). As
to personality disorders, however, less is known. Some
studies have been performed, but few of them adequately represent the general population (Torgersen et
al. 2001). In this chapter I review published studies that
are closest to what one might call an epidemiological
population study. These individual studies are discussed below in view of different elements of epidemiology, beginning with a discussion of sample selection
for each study.

Sample Selection
The sample studied by Zimmerman and Coryell (1989,
1990) included first-degree relatives of normal subjects
(23%) and of psychiatric patients (mood disorders and
schizophrenia) as well as a smaller group of first-degree
relatives of nonpsychotic psychiatric patients. Thus,
even if this is a nonpatient sample it is not an average
population sample. However, the prevalence of mania
was not higher than 2%, and the prevalence of schizophrenia was not higher than 1%. Twenty-seven percent
129

130

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

of the interviews were conducted in person and the remainder by telephone. The Structured Interview for
DSM-III Personality Disorders was applied (Stangl et
al. 1985). The study took place in Iowa City.
The sample reported on by Black et al. (1993) consisted of 120 relatives of 32 outpatients with obsessivecompulsive disorder and 127 relatives of a comparison
group screened for Axis I disorders. Strangely, no difference was found between the prevalence of personality disorders in the two relative groups. More than half
of the sample were siblings, a quarter were parents,
and the rest were children. A little more than half were
women. The mean age was 42 years. More than half
were interviewed in person and the rest by telephone.
Maier et al. (1992) conducted one of the few reported studies in which the sample is relatively representative of the general population. Control probands
were selected by a marketing company to match patients older than age 20 on sex, age, residential area, and
educational level. The participants had to have at least
one living first-degree relative who also had agreed to
be interviewed. Otherwise, this sample represented the
general population of a mixed urban/rural German
residential area near Mainz. No screening for medical
or psychiatric history was performed. The control
probands, their spouses, and first-degree relatives constituted the sample.
The sample studied by Moldin et al. (1994) consisted
of parents and their offspring in two control groups used
in the New York High Risk Project. One of the groups
was recruited from two schools in the New York metropolitan area. The other group came from the pool of a
population sampling firm. The subjects were white, English-speaking families screened for psychiatric disorder.
In the study by Klein et al. (1995), the sample comprised relatives of a control group screened for Axis I disorders in Stony Brook, New York. The interviews were
partly conducted in person and partly by telephone.
Lenzenweger et al. (1997) examined a sample consisting initially of 1,684 undergraduate students from
Cornell University in New York. They were screened
by means of a questionnaire; a sample of those expected and those not expected to have a personality
disorder was interviewed. The total number of subjects interviewed was 258. In this overview (Table 81),
I apply the actual numbers. The estimated prevalence
for any personality disorder is a little different.
The study by Torgersen et al. (2001) was conducted
in Oslo, the capital of Norway. A random sample of
names of 3,590 citizens between 18 and 65 years of age
was selected from the National Register of Oslo. Some
had moved out of town, some were impossible to trace,

and some were dead. Others refused to participate or


postponed the interview beyond the period of the study
(18%). Of the original sample, 2,053 (57%) delivered interviews of sufficient quality for the study. All interviews were performed in person. The sampling procedure made it possible to identify all causes of reduction
in the sample from the initial to the final sample. There
were almost equal numbers of men and women.
The sampling procedure used by Samuels et al.
(2002) was very complicated. Initially, a sample of 3,481
adult household residents in Baltimore was studied in
the 1980s. About 10 years later, a subsample was selected that included individuals previously evaluated
by psychiatrists or those who appeared to have an Axis
I diagnosis based on the Diagnostic Interview Schedule.
In addition, a random sample was selected. A number
of subjects could not be traced, refused, were too ill to
participate, or were deceased. The remaining sample
consisted of 742 individuals. Their ages varied between
34 and 94 years, and two-thirds were women.

Results
Table 81 presents the prevalences in the published
studies discussed above, including all personality disorders. So-called mixed personality disorders, defined
by the absence of one criterion for two or more personality disorders and not having the required number of
criteria for any disorder, are excluded. Unweighted
prevalences (rather than weighted prevalences based
on questionable weighting procedures) are presented,
because the prevalences among those not reached cannot be known. The qualified, although questionable,
guesswork gives one an impression of increased accuracy. A nonweighted rate is transparent and does not
claim more than it can stand for.
The prevalence of any personality disorder varies
between 3.9% and 22.7%. If the small samples of 303
and under are disregarded, the variation is much less,
from 10.0% to 14.3%. The median prevalence of all the
studies for any personality disorder is 11.55%, and the
pooled prevalence is 12.26%.
As to the specific disorders, the prevalence of obsessive-compulsive and passive-aggressive personality disorders is around 2%, regardless of whether the
median or pooled prevalence is used. For avoidant
personality disorder, the result for median and pooled
prevalence is somewhat different (1.23% vs. 2.92%) because of the high prevalence in the large Norwegian
study (Torgersen et al. 2001), perhaps consistent with
the low genetic loading of avoidant personality disorder in Norway (Torgersen et al. 2000) (see Chapter 9,

Table 81.

Prevalences of personality disorders in eight epidemiological studies

Place
Method
System

Zimmerman
and Coryell
1989

Black et al.
1992

Iowa
SIDP
DSM-III

Iowa
SIDP
DSM-III

Mainz
SCID-II
DSM-III-R

NYC
PDE
DSM-III-R

New York
PDE
DSM-III-R

New York
PDE
DSM-III-R

Oslo
SIDP-R
DSM-III-R

Baltimore
IPDE
DSM-IV

0.9
0.9
2.9
3.3
1.7
3.0
0.0
1.3
1.8
2.0

1.6
0.0
3.2
0.8
3.2
3.2
0.0
2.0
1.6
9.3

1.8
0.4
0.7
0.2
1.1
1.3
0.0
1.1
1.6
2.2

0.0
0.0
0.7
2.6
2.0
0.3
0.0
0.7
1.0
0.7

1.8
0.9
0.0
2.6
1.8
1.8
4.4
5.7
0.4
3.1

0.4
0.4
0.0
0.8
0.0
1.9
1.2
0.4
0.4
0.0

2.2
1.6
0.6
0.6
0.7
1.9
0.8
5.0
1.5
1.9

0.7
0.7
1.8
4.5
1.2
0.4
0.1
1.4
0.3
1.2

3.3

14.3

10.5

22.7

1.8

10.0

1.7

7.3

1.8

14.8

0.0
0.0
0.0
3.9

1.6
0.8
0.2
13.1

10.0

Maier et al. Moldin et al. Klein et al. Lenzenweger Torgersen et al. Samuels et al.
1992
1994
1995
et al. 1997
2001
2002

Range

Median

Pooled

0.02.2
0.01.6
0.03.2
0.24.5
0.03.2
0.43.2
0.04.4
0.45.0
0.41.8
0.09.3

1.25
0.65
0.70
1.70
1.45
1.85
0.05
1.35
1.30
1.95

1.48
0.96
1.20
1.77
1.16
1.77
0.61
2.91
1.24
2.09

0.010.5
0.00.83
0.00.19
3.922.7

1.80
0.40
0.10
11.55

1.99
0.74
0.17
12.26

5,081

5,081

Personality disorder

Number

797

247

452

303

229

258

2,053

742

Epidemiology

Paranoid
Schizoid
Schizotypal
Antisocial
Borderline
Histrionic
Narcissistic
Avoidant
Dependent
Obsessivecompulsive
Passive-aggressive
Self-defeating
Sadistic
Any personality
disorder

Note. IPDE=International Personality Disorders Examination; NYC=New York City; PDE=Personality Disorder Examination; SCID-II=Structured Clinical Interview for DSM-IV Axis II Personality Disorders; SIDP= The Structured Interview for DSM-III Personality Disorders; SIDP-R=The Structured Interview for DSM-III-R Personality Disorders.

131

132

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Genetics). If there is cultural pressure in the direction of avoidant behavior, then the prevalence will increase and the genetic estimate will decrease as the
environmental estimate increases. However, three of
eight studies have found a prevalence above 2%, so
the true prevalence is probably not much lower. The
prevalence of histrionic and antisocial personality disorder is between 1.5% and 2%.
The prevalence of paranoid, dependent, and borderline personality disorder seems to be between 1%
and 1.5%. As to schizotypal personality disorder, the
difference between median and pooled prevalence is
mostly due to large variation among the studies. However, an estimate of 1% may be quite good. Finally, the
prevalence of schizoid and self-defeating personality
disorder is between 0.5% and 1%, and the prevalence
of narcissistic and sadistic personality disorder appears to be even smaller.
Table 82 shows a comparison between the prevalences in a large outpatient clinic in Oslo (Alns and
Torgersen 1988) and in the general population of that
city (Torgersen et al. 2001). The ratio between the prevalence in the clinic and that in the population is calculated separately for women and men and in the total
sample. There are relatively small differences in the
range of the ratios between women and men, even if
the ratios are a little larger for the specific personality
disorder among men (not for all personality disorders).
Those with dependent, borderline, avoidant, and obsessive-compulsive personality disorder are strongly
overrepresented among the patients based on prevalence rates in the general population, whereas those
with antisocial, schizoid, and paranoid personality disorder are less common in the clinical compared with
the general population. To have a borderline, avoidant,
or schizotypal personality disorder implies pain and
dysfunction, as I discuss later in the chapter. One may
speculate that those who are dependent seek help,
whereas obsessive-compulsive patients want to do
something with their problems, even if they do not suffer as much. In the other direction, those who are antisocial do not want psychological help and are also refused help. Schizoid individuals keep their distance,
whereas paranoid subjects do not believe in any cure.

SOCIODEMOGRAPHIC CORRELATES
Gender
Gender differences are common among mental disorders. Women more often have mood and anxiety disor-

ders, and men more often have substance-related disorders (Kringlen et al. 2001). For personality disorders,
women and men also differ (Chapter 34, Gender).
With regard to personality disorders, Zimmerman
and Coryell (1989) observed a higher prevalence of
personality disorders among males, as did Jackson
and Burgess (2000) for ICD-10 screening when regression analysis was applied. However, differences between genders were very small, and Torgersen et al.
(2001) did not observe any differences.
As to the personality disorder clusters, Samuels et
al. (2002) and Torgersen et al. (2001) reported that
Cluster A (odd/eccentric) and Cluster B (dramatic/
emotional) personality disorders or traits were more
common among men. Among the specific Cluster A
disorders, both Torgersen et al. (2001) and Zimmerman
and Coryell (1990) found that schizoid personality disorder or traits were more common among men. Zimmerman and Coryell (1990) found this also for paranoid traits, and neither Zimmerman and Coryell (1989,
1990) nor Torgersen et al. (2001) observed any difference for schizotypal personality disorder. Among the
Cluster B personality disorders, antisocial disorder is
much more common among men (Torgersen et al.
2001; Zimmerman and Coryell 1989, 1990). Those with
histrionic personality disorder or traits appear more often to be women (Torgersen et al. 2001; Zimmerman
and Coryell 1990). Narcissistic traits are found more often among men, and there are no statistically significant gender differences for borderline personality disorder or traits (Torgersen et al. 2001; Zimmerman and
Coryell 1990). Among the Cluster C (anxious/fearful)
personality disorders, dependent personality disorder
is much more common among women, and obsessivecompulsive personality disorder or traits are found
more often among men (Torgersen et al. 2001; Zimmerman and Coryell 1989, 1990); only Zimmerman and
Coryell (1989, 1990) reported more avoidant personality disorder and traits among women.
Regarding personality disorders provided for further study (American Psychiatric Association 2000),
Torgersen et al. (2001)(but not Zimmerman and
Coryell 1989, 1990)found that men more often had
passive-aggressive personality disorder. Torgersen
and colleagues also found that women more often presented with self-defeating traits, and men more often
presented with sadistic traits.
The most clear-cut results from the studies are that
men tend to be antisocial and women tend to be dependent. These results are perhaps not surprising.
However, more surprising is a lack of gender difference for borderline traits; in patient samples border-

Table 82.

Prevalences of personality disorders in the common population and among outpatients in Oslo, Norway
Females

Personality
disorder

Males

Total

Torgersen et
al. 2001

Alns and
Torgersen 1988

Ratio
(range)

Paranoid

2.2

3.9

1.8 (9)

Schizoid

1.1

0.0

0.0 (10)

2.2

5.4

Schizotypal

0.6

3.9

6.5 (6)

0.5

12.0

Antisocial

0.0

0.0

0.0 (10)

1.3

0.0

Borderline

0.9

17.0

0.4

9.8

Histrionic

2.5

15.0

6.0 (7)

1.2

Narcissistic

0.8

1.9

2.4 (8)

0.9

Avoidant

5.0

53.4

10.7 (3)

4.9

59.8

12.2 (5)

5.0

55.4

11.1 (3)

Dependent

2.0

47.6

23.8 (1)

0.9

45.7

50.8 (1)

1.5

47.0

31.3 (1)

Obsessivecompulsive

1.3

13.6

10.5 (4)

2.6

33.7

13.0 (4)

1.9

19.8

10.4 (5)

Passiveaggressive

0.9

6.3

7.0 (5)

2.2

18.5

8.4 (8)

1.6

10.1

6.3 (7)

12.6

76.7

13.7

90.2

6.6

13.1

80.9

6.2

Number

1,142

206

6.1

2.3

911

7.6

Ratio
(range)

Torgersen et al.
Alns and
2001
Torgersen 1988

3.3 (9)

2.2

5.0

2.5 (10)

1.6

1.7

0.6

6.4

24.0 (3)

2.3 (9)
1.1 (10)
10.7 (4)

0.6

0.0

24.5 (2)

0.7

14.8

21.1 (2)

10.9

9.1 (7)

1.9

13.8

7.3 (6)

10.9

12.1 (6)

0.8

4.7

5.9 (8)

92

0.0 (11)

Ratio
(range)

2,053

298

0.0 (11)

Epidemiology

Any personality
disorder

18.9 (2)

Torgersen et al.
Alns and
2001
Torgersen 1988

133

134

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

line personality disorder is not more prevalent among


women than among men (Alns and Torgersen 1988;
Fossati et al. 2003; Golomb et al. 1995). In one study
borderline personality disorder was, in fact, more
common among men than among women (Carter et
al. 1999). In our unsystematic impression of people,
we are more likely to see borderline features in
women than in men. That schizotypal personality disorder does not show any gender bias will more easily
be recognized. A trend in the direction of men being
schizoid, narcissistic, and obsessive-compulsive and
women being more histrionic is in accordance with
common opinion.

Age
To diagnose a personality disorder in an individual under the age of 18 years, the features must have been
present at least 1 year (American Psychiatric Association 2000). At the same time, it is assumed that personality disorders start early in life and are relatively stable. For some personality disorders, especially the
dramatic types, it is also assumed that they are typical
for young people. On the other hand, the older people
are, the longer they have had to develop personality
disorders, even though personality disorders may also
disappear. Suicide and fatal accidents also may happen
more often among those with personality disorders
than among other individuals. This fact will influence
the rate of specific personality disorders in older age.
What does empirical research show? Zimmerman
and Coryell (1989) observed that those with personality disorders were younger than those without. Jackson and Burgess (2000) found the same using a short
ICD-10 screening instrument (International Personality Disorders Examination screener). Torgersen et al.
(2001), however, observed the opposite. This opposite
finding can be explained by the high prevalence of introverted and low prevalence of impulsive personality
traits in Norway as compared with the United States.
As to the clusters of personality disorders, Torgersen et al. (2001) found that individuals with odd/
eccentric personality disorders were older, whereas
Samuels et al. (2002) did not find any age variation. For
the dramatic/emotional cluster, Samuels et al. (2002)
found a higher prevalence among the younger subjects, whereas Torgersen et al. (2001) found that the
dramatic/emotional trait dimensions decreased with
age. As to the anxious/fearful cluster, neither group
observed any age trend.
Among the odd/eccentric personality disorders,
schizoid personality disorder or traits seem to be asso-

ciated with being older (Torgersen et al. 2001; Zimmerman and Coryell 1989, 1990). Paranoid personality
disorder is unrelated to age (Torgersen et al. 2001;
Zimmerman and Coryell 1989, 1990), whereas Zimmerman and Coryell (1989, 1990) observed that those
with schizotypal personality disorder were younger,
and Torgersen et al. (2001) found that they were older.
Among the dramatic/emotional personality disorders, those with antisocial and borderline personality
disorder or traits are younger (Torgersen et al. 2001;
Zimmerman and Coryell 1989, 1990), and Zimmerman
and Coryell (1990) observed that those with histrionic
and narcissistic traits are younger as well. These results
are not confirmed by Torgersen et al. (2001).
Zimmerman and Coryell (1989, 1990) did not find
any age trend for any of the fearful disorders, whereas
Torgersen et al. (2001) observed that individuals with
obsessive-compulsive disorder and avoidant traits are
older. No difference was found for dependent personality disorders.
Among the proposed personality disorders, Zimmerman and Coryell (1989) found that those with passive-aggressive personality disorder are younger, and
Torgersen et al. (2001) observed also that such traits
were negatively correlated with age. The latter study
also examined self-defeating and sadistic traits and
found that sadistic traits were associated with being
younger.
To summarize, persons with schizoid personality
disorder appear to be older, and persons with antisocial and borderline personality disorder seem to be
younger. Perhaps individuals with obsessive-compulsive and avoidant disorders also are older, and those
with histrionic and narcissistic disorders are younger.
The reason for this age difference in disorders is that
people become more introverted and obsessive and
less impulsive and overtly aggressive as they age.
Thus, the basic relative frequency of odd/eccentric
and anxious/fearful versus dramatic/emotional personality disorders in a population will determine
whether having any personality disorder is more frequent in younger or older age.

Marital Status
Most of the results concerning marital status are
from Zimmerman and Coryell (1989). Some of the
data from Torgersen et al. (2001) have been calculated
for this chapter to fit the tables in Zimmerman and
Coryell (1989) (see Table 83).
As illustrated in Table 83, subjects with personality disorder have more often been separated or di-

Table 83.

Marital status and personality disorders, calculated from Torgersen et al. (2001)

Number

Single
(never
married)

Married

Separated

Divorced

Widowed

Paranoid

46

34.8

34.8

6.5

21.7a

2.2

15.8

36.7

Schizoid

32

56.3

31.3

0.0

6.3

6.3

20.0

28.6

Schizotypal

12

50.0

33.3

0.0

8.3

8.3

20.0

16.7

Personality disorder

8.3a

Ever separatede

Ever divorcedd

Antisocial

12

75.0

0.0

16.7

0.0

0.0

66.7

Borderline

14

57.1

35.7

7.1

0.0

0.0

20.0

16.7

Histrionic

39

46.2

35.9

0.0

17.9

0.0

0.0

47.6a

Narcissistic

17

35.6

52.9

0.0

5.9

5.9

10.0

9.1

Avoidant

45.1

36.3

1.0

14.7

2.9

7.5

28.6

31

58.1a

25.8a

3.2

12.9

0.0

11.1

30.8

Obsessive-compulsive

39

41.6

43.6

0.0

10.3

5.1

5.6

21.7

Passive-aggressive

32

35.3

31.3

6.3

9.4

3.1

18.2

31.3

0.0

41.2c

5.9

25.0

63.6

0.0

0.0

0.0

0.0

0.0

3.8

15.0

2.5

13.8

34.1

1.4

12.7

1.4

8.3

33.3

Self-defeating
Sadistic

17

35.3

17.6

50.0

56.0

Eccentric

80

45.6

33.8

Dramatic

62

49.3

35.2

Fearful

189

45.5

36.5

1.3

14.1

2.6

8.2

28.2

Any personality disorder

269

43.9

36.8b

2.2

15.6a

1.5

7.9

33.1b

No personality disorder

1,784

38.8

46.5

2.4

10.4

1.8

5.1

23.2

Number

2,053

693

830

43

185

33

43

Epidemiology

102

Dependent

253

aX2 test,

P<0.05
X test, P<0.01
c 2
X test, P <0.001
dExcluding those who are never married
eExcluding those who are never married, and those who are divorced
b 2

135

136

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

vorced compared with those without a personality


disorder, and they are more often divorced at the time
of the interview (Zimmerman and Coryell 1989). They
are less frequently married (Jackson and Burgess 2000;
Zimmerman and Coryell 1989), and they are more often never married (Zimmerman and Coryell 1989). If
we include living nonmarried persons with a partner,
subjects with personality disorder live more often
alone without a partner compared with those without
a personality disorder in the general population (Torgersen et al. 2001).
However, as the risk of having a personality disorder is related to gender and age, the real effect of other
sociodemographic variables such as marital status is
difficult to determine. Younger people are less often
married, and education is related to gender and age.
The best way to determine the independent effect of
other sociodemographic variables is to apply multivariate methods. However, to apply such methods one
needs large samples. Thus multivariate methods have
been used in very few studies. In the study of Torgersen et al. (2001), such multivariate analyses have
been carried out for living alone versus living with a
partner.
Those with eccentric personality disorders have
more often been divorced or separated, they are more
often divorced when interviewed, and are seldom
married (Samuels et al. 2002) (Table 83). Those with
dramatic personality disorders are also often unmarried and live more often alone (Torgersen et al. 2001).
Those with fearful personality disorders are also less
often married (Samuels et al. 2002) and live more often
alone (Torgersen et al. 2001).
When we examine the specific personality disorders, we find little correspondence between the studies by Zimmerman and Coryell (1989) and Torgersen
et al. (2001). Marital status does not seem to be as important in the Norwegian study, perhaps because
many Norwegians live in stable relationships without
being married. When we include living together with
a partner from the study of Torgersen et al. (2001) and
consider this life situation as analogous to marriage,
we find more similarity between the two studies. It is
important to note that the relationships in the Torgersen et al. study are based on logistic and linear regression analysis, taking into account a number of
other sociodemographic variables.
Among the odd/eccentric personality disorders,
those with paranoid personality disorder are more often divorced (Table 83) and living alone (Torgersen et
al. 2001). Those with schizoid personality disorder are
more seldom separated (Zimmerman and Coryell

1989), less often married (Torgersen et al. 2001), and


more often living alone (Torgersen et al. 2001). Those
with schizotypal personality disorder have more often
been separated (Zimmerman and Coryell 1989) and
live more often alone (Torgersen et al. 2001).
The only personality disorder for which the American and Norwegian studies have reached the same
conclusion is for one of the dramatic/emotional personality disorders, namely histrionic personality disorder. Persons with histrionic personality disorder
have more often been separated or divorced (Zimmerman and Coryell 1989). They are also more often not
married when interviewed (Zimmerman and Coryell
1989) and live more often alone (Torgersen et al. 2001).
Those with antisocial personality disorder have also
more often been divorced or separated (Zimmerman
and Coryell 1989), are less often married when interviewed, and live more often alone (Torgersen et al.
2001). Persons with borderline personality disorder
also have more often been separated if married, are
more often divorced, and are not married when interviewed (Zimmerman and Coryell 1989). They are
more often never married, and live more often alone
(Torgersen et al. 2001). Finally, those with narcissistic
personality disorder also more often live alone (Torgersen et al. 2001).
Among persons with anxious/fearful personality
disorders, those with avoidant personality disorder
and dependent personality disorder have more often
been separated (Zimmerman and Coryell 1989). They
are more often separated (Zimmerman and Coryell
1989) and not married when interviewed and live more
often alone (Torgersen et al. 2001). Those with obsessive-compulsive traits are less often married (Torgersen
et al. 2001).
Among the proposed personality disorders, persons with passive-aggressive personality disorder
have more often been divorced and are less often
married when interviewed (Zimmerman and Coryell
1989) and live more often alone (Torgersen et al. 2001).
Those with self-defeating personality disorder have
more often been divorced, are more often divorced
and not married when interviewed, and live more often alone (Torgersen et al. 2001).

Education and Income


Very few studies have investigated the relationship
between personality disorders and education and income. Torgersen et al. (2001) observed that those with
any personality disorder had less education. The same
was observed for those with odd/eccentric personal-

Epidemiology

ity disorders, and those with dramatic/emotional as


well as anxious/fearful personality disorder traits.
Samuels et al. (2002) confirmed that those with dramatic/emotional personality disorders had less education but not those with odd/eccentric or anxious/
fearful personality disorders.
As to the specific personality disorders, results are
published only from the study of Torgersen et al. (2001).
They observed, applying logistic regression analysis
and taking into account a number of other sociodemographic variables, that only those with paranoid and
avoidant personality disorders had less education than
those without the disorders. Interestingly, those with
obsessive-compulsive disorder in fact had higher education than those without the disorder.
If the personality disorders are treated as continuous variables by adding traits, we find that all personality disorder traits, except histrionic, narcissistic, and
passive-aggressive, are related to education by applying linear regression analysis. All are negatively related except obsessive-compulsive, which is positively
related.
Samuels et al. (2002) also investigated the relationship between income and personality disorders but
did not find any association. Jackson and Burgess
(2000) did not find any relationship to unemployment.
It is important to note that all of these studies applied
multivariate methods while taking into account other
sociodemographic variables.

Urban Location
The study of Torgersen et al. (2001) showed that those
living in the populated center of the city more often
had a personality disorder. The same was true for all
clusters of personality disorders and all specific disorders except antisocial, sadistic, avoidant, and dependent personality.
One may speculate about why there are more personality disorders in the center than in the outskirts of
the city. Quality of life is generally lower in the center
of the city (Cramer et al. 2004), and there is a higher
rate of symptom disorders in the city or in the center
of the city (Kringlen et al. 2001; Lewis and Booth 1992,
1994; Marcelis et al. 1998; Sundquist et al. 2004; van Os
et al. 2001). One reason may be that the concentrated
urban life creates stress leading to personality disorders. Another reason may be that individuals with
personality problems drift to the center, where they
can lead an anonymous life. A third explanation may
be that less social control simply makes it easier to express the less socially acceptable aspects of ones per-

137

sonality. We used to think that excessive social control


creates mental problems. Perhaps social control hinders the development of problemsnot only aggressive, antisocial personality styles but also accentuated
eccentric, narcissistic, impulsive, and extraverted personality styles.

QUALITY OF LIFE AND DYSFUNCTION


Central to the definition of personality disorder are the
interpersonal problems, reduced well-being, and dysfunction that personality disorders imply. Only one
study has investigated reduced quality of life among
those with personality disorders (Cramer et al. 2003). In
the sample studied by Torgersen et al. (2001), quality of
life was assessed by interview and included the following aspects: subjective well-being, self-realization, relation to friends, social support, negative life events, relation to family of origin, and neighborhood quality. All
aspects were integrated in a global quality-of-life index.
The results showed that among the odd/eccentric
personality disorders, schizotypal personality disorder
implied the most reduced quality of life, followed by
schizoid and paranoid personality disorders. Among
the dramatic/emotional personality disorders, those
with borderline personality disorder had the most reduced quality of life, followed by narcissistic and antisocial personality disorders. Those with histrionic personality disorder had only a slightly reduced quality of
life. Among the anxious/fearful personality disorders,
those with avoidant personality disorder had the lowest quality of life, followed by dependent personality
disorder. Those with obsessive-compulsive personality disorder had only slightly reduced quality of life.
Among the personality disorders provided for further study, those with self-defeating personality disorder (American Psychiatric Association 1987) had
somewhat reduced quality of life, whereas those with
passive-aggressive personality disorder (American
Psychiatric Association 2000) had only slightly reduced quality of life.
A dysfunction index was created by combining
quality of life (reversed); the answer to the Structured
Interview for DSM-III Personality DisordersRevised
question do you feel that the way you usually deal
with people and handle situations causes you problems?; the number of lifetime Axis I diagnoses; and
any incidence of seeking treatment with varying degrees of seriousness, from private psychologist and
psychiatristvia outpatient and inpatient clinicsto
psychiatric hospitals. The dysfunction index was re-

138

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

2.45

2.36

2.5
2.09
2

1.76

Quality of life

1.48

1.37

1.5
0.95
1
0.51
0.5

0
0

79

Number of criteria

Figure 81. The relationship between maximum number of criteria fulfilled on any personality disorder and
quality of life.
As explained in text, the ordinate (quality of life) is a composite of subjective well-being, self-realization, social support, lack of negative lifeevents, and relation to family, friends and neighbors. The mean is set to 2 and the standard deviation is 1.

lated to personality disorder much as the global quality-of-life index was. The only differences found in
comparing results derived from the dysfunction index
with those from the global quality-of-life index for
dramatic/emotional personality disorders were that
those persons with borderline and histrionic personality disorder appeared more dysfunctional, and those
persons with antisocial personality disorder appeared
less dysfunctional. Among the anxious/fearful personality disorders, those persons with dependent personality disorder became more similar to those with
avoidant personality disorder, and among the proposed personality disorders the difference between
self-defeating and passive-aggressive personality disorders increased strongly. The reason for the differences is mainly that those with borderline, histrionic,
dependent, and self-defeating personality disorders
are more likely to seek treatment and those with antisocial and passive-aggressive personality disorders
are less likely to seek treatment.
However, the most important result in this study
was that for both quality of life and dysfunction, there

was a perfect linear doseresponse relationship to


numbers of criteria fulfilled for all personality disorders together and to the number of criteria fulfilled for
any specific personality disorder. Thus, if a person has
one criterion fulfilled for any personality disorder, the
quality of life is lower and dysfunction is higher than
among those with no criteria fulfilled. Those with two
criteria fulfilled on any disorder or a specific disorder
have more problems than those with one, those with
three criteria have more problems than those with
two, and so on. When those with zero criteria on all
disorders are grouped togetherthat is, those with a
maximum of one criterion on any disorder, those with
a maximum of two, and so onthe relationship to global quality of life and dysfunction was perfectly linear
(Figures 81 and 82). This result means that there are
no arguments for any specific number of criteria to define a personality disorder if one uses quality of life or
dysfunction as validation variables. There is no natural cutoff point.
A high level of dysfunction and disability was also
observed among those with schizotypal, borderline,

139

Epidemiology

2.69

2.5
Level of dysfunction

2.05
2
1.59
1.5

1.73

1.25
0.91

1
0.47

0.62

0.5
0
0

7 9

Number of criteria
Figure 82. The relationship between maximum number of criteria fulfilled on any personality disorder
and dysfunction.
As explained in text, the ordinate (dysfunction) is a composite of life quality (reversed), treatment seeking, comorbid Axis I disorders, and
the notion that ones behavior causes problems. The mean and standard deviation is 1.

and avoidant personality disorders in a large-scale


multicenter study of patients with personality disorders (Skodol et al. 2002). It was also observed that
those with obsessive-compulsive personality disorder
showed much less disability.
Zimmerman and Coryell (1989) also found a high
frequency of psychosexual dysfunction among persons with avoidant personality disorder. Surprisingly,
this dysfunction was infrequent for persons with
borderline personality disorder, and, not surprisingly,
it was also infrequent for those with antisocial personality disorder.
In the future, there is reason to believe that we will
see more studies of quality of life, dysfunction, and
disability among subjects with personality disorders,
either in the general population or in patient samples.

CONCLUSIONS
Personality disorders are prevalent: more than 1 in 10
adult individuals has a personality disorder.
The average prevalence of the specific personality
disorders is a little above 1%, somewhat higher for ob-

sessive-compulsive, passive-aggressive, avoidant, histrionic, and antisocial personality disorders, and somewhat lower for sadistic, narcissistic, and self-defeating
personality disorders. Two of these low-prevalence
disorders, sadistic and self-defeating, are only provided for further study in DSM-III-R (American Psychiatric Association 1987), and none of the three exists
in ICD-10.
Those with dependent, borderline, obsessive-compulsive, avoidant, and schizotypal personality disorders are overrepresented in patient populations, both
overall and when gender is controlled for, whereas
those with antisocial, schizoid, and paranoid personality disorders are underrepresented.
The highest prevalences of personality disorders in
the general population are observed among subjects
with lower education living in populated areas, for example in the city center. They often have a history of
divorce and separation and are more often living without a partner.
Men have typically a schizoid, antisocial, or obsessive-compulsive personality disorder, whereas women
are more inclined toward a dependent or histrionic personality disorder. Antisocial, borderline, and passive-

140

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

aggressive personality disorders are more often observed among younger persons, whereas older individuals more often have a schizoid personality disorder.
Typical for those with histrionic personality disorder is
to have a history of divorces and separations. These individualstogether with those with other dramatic/
emotional personality disorders, such as borderline and
antisocial personality disordersare often unmarried,
divorced, or generally live alone without a partner. The
same is also true for those with passive-aggressive personality disorder.
Lower education is most typical for those with
paranoid and avoidant personality disorders, whereas
those with obsessive-compulsive personality disorder
in fact have higher education than those without the
disorder. Those with paranoid, schizotypal, histrionic,
and passive-aggressive personality disorders are most
inclined to live in the city center.
Personality disorders imply dysfunction and reduction in quality of life, including reduced subjective wellbeing and self-realization, relational problems, lack of
social support, and frequent negative life events.
Among the personality disorders, individuals with
schizotypal, borderline, and avoidant personality disorders tend to have the most reduced quality of life,
dysfunction, and disability. Individuals with obsessivecompulsive, histrionic, and passive-aggressive personality disorders tend to have the least reduction in quality of life, dysfunction, and disability.
There is an even reduction in quality of life and an
even increase in dysfunction for each criterion manifested. Thus, there is a continuous relationship between those with no or small personality problems,
those with moderate problems, and those with severe
problems. No natural cutoff point exists. Any definition of how many criteria are required for a personality disorder is arbitrary. Even so, to have a definition is
important for communication. However, a change in
criteria will immediately change the prevalence estimates in the society. Consequently, correlations between personality disorders and other variables are
more important than prevalence rates. These correlations appear to be independent of how strictly personality disorders are defined.
Epidemiological research has perhaps changed
some stereotypic notions about personality disorders.
Personality disorders are more frequent in the general
population than we generally believed, especially the
introverted personality disorders. Borderline personality disorder is not a female disorder. Living without
a partner is a risk factor for personality disorders, but
being unmarried is less a risk factor than many would

have believed. Those living in a partnership without


being married function well.
Care must be taken to avoid believing that these correlations display one-directional causal relationships.
Personality disorders may hinder obtaining higher levels of education and may be created by socioeconomic
difficulties. Problematic personality traits may prevent
a person from going into a relationship or may lead to
the breaking-up of relationships, rather than having relationship issues and problems causing problematic
personality traits.
Personality disorders are not something that a person has for life. Impulsive, aggressive, and extraverted
features may decrease with age, whereas obsessiveintroverted traits increase as people get older.
Perhaps one of the most important aspects of personality disorders is the reduction of quality of life
that is implied. However, a low quality of life does not
necessarily create a personality disorder; the opposite
is just as likely.

REFERENCES
Alns R, Torgersen S: DSM-III symptom disorders (Axis I)
and personality disorders (Axis II) in an outpatient population. Acta Psychiatr Scand 78:348355, 1988
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition. Washington,
DC, American Psychiatric Association, 1980
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition, Revised. Washington, DC, American Psychiatric Association, 1987
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition. Washington,
DC, American Psychiatric Association, 1994
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association, 2000
Black DW, Noyes R Jr, Pfohl B, et al: Personality disorder in
obsessive-compulsive volunteers, well comparison subjects, and their first-degree relatives. Am J Psychiatry
150:12261232, 1993
Carter JD, Joyce PR, Mulder RT, et al: Gender differences in
the frequency of personality disorders in depressed outpatients. J Personal Disord 13:6774, 1999
Cramer V, Torgersen S, Kringlen E: Personality disorders,
prevalence, sociodemographic correlations, quality of
life, dysfunction, and the question of continuity. Persnlichkeitsstrungen. Theorie und Therapie 7:189198,
2003
Cramer V, Torgersen S, Kringlen E: Quality of life in a city:
the effect of population density. Social Indicator Research 69:103116, 2004

Epidemiology

Fossati A, Feeney JA, Donati D, et al: Personality disorders


and adult attachment dimensions in a mixed psychiatric
sample: a multivariate study. J Nerv Ment Dis 191:3037,
2003
Golomb M, Fava M, Abraham M, et al: Gender differences in
personality disorders. Am J Psychiatry 152:579582, 1995
Jackson HP, Burgess PM: Personality disorders in the community: a report from the Australian National Survey of
Mental Health and Well Being. Soc Psychiatry Psychiatr
Epidemiol 35:531538, 2000
Klein DN, Riso LP, Donaldson SK, et al: Family study of
early onset dysthymia: mood and personality disorders
in relatives of outpatients with dysthymia and episodic
major depressive and normal controls. Arch Gen Psychiatry 52:487496, 1995
Kringlen E, Torgersen S, Cramer V: A Norwegian psychiatric
epidemiological study. Am J Psychiatry 158:10911098,
2001
Lenzenweger MF, Loranger AW, Korfine L, et al: Detecting
personality disorders in a nonclinical population: application of a 2-stage procedure for case identification.
Arch Gen Psychiatry 54:345351, 1997
Lewis G, Booth M: Regional differences in mental health in
Great Britain. J Epidemiol Community Health 46:608
611, 1992
Lewis G, Booth M: Are cities bad for your mental health?
Psychol Med 24:913915, 1994
Maier W, Lichtermann D, Klingler T, et al: Prevalences of
personality disorders (DSM-III-R) in the community.
J Personal Disord 6:187196, 1992
Marcelis M, Navarro-Mateu F, Murray R: Urbanization and
psychosis: a study of 19421978 birth cohorts in the
Netherlands. Psychol Med 28:11971203, 1998

141

Moldin SO, Rice JP, Erlenmeyer-Kimling L, et al: Latent


structure of DSM-III-R Axis II psychopathology in a
normal sample. J Abnorm Psychol 103:259266, 1994
Samuels J, Eaton WW, Bienvenu OJ III, et al: Prevalences and
correlates of personality disorders in a community sample. Br J Psychiatry 180:536542, 2002
Skodol AE, Gunderson JG, McGlashan TH, et al: Functional
impairment in patients with schizotypal, borderline,
avoidant, or obsessive-compulsive personality disorder. Am J Psychiatry 159:276283, 2002
Stangl D, Pfohl B, Zimmerman M, et al: A structured interview for the DSM-III personality disorders. Arch Gen
Psychiatry 42:591596, 1985
Sundquist K, Frank G, Sundquist J: Urbanisation and incidence of psychosis and depression. Br J Psychiatry
184:293298, 2004
Torgersen S, Kringlen E, Cramer V: The prevalence of personality disorders in a community sample. Arch Gen
Psychiatry 58:590596, 2001
Torgersen S, Lygren S, ien PA, et al: A twin study of personality disorders. Compr Psychiatry 41:416425, 2000
van Os J, Hanssen M, Bijl RV, et al: Prevalence of psychotic disorder and community level of psychotic symptoms: an
urban-rural comparison. Arch Gen Psychiatry 58:663
668, 2001
Zimmerman M, Coryell W: DSM-III personality disorder diagnoses in a nonpatient sample: demographic correlates and comorbidity. Arch Gen Psychiatry 46:682689,
1989
Zimmerman M, Coryell WH: Diagnosing personality disorders in the community: a comparison of self-report and
interview measures. Arch Gen Psychiatry 47:527531,
1990

This page intentionally left blank

9
Genetics
C. Robert Cloninger, M.D.

dimensions of personality. Essentially the same dimensions of personality are observed whether one begins
with normal personality variation in the general community or with symptoms of personality disorder in
treatment samples, as described in detail in the next
section. Different investigators have described personality traits with differing content depending on their
methods and understanding of personality development, but the alternative systems overlap extensively
(Cloninger et al. 1994; Livesley et al. 1998; Svrakic et al.
1993; Zuckerman and Cloninger 1996). Fine-grained
descriptions are based on 2030 subscales and higherorder descriptions are based on 57 scales resulting
from the grouping of the subscales. One comprehensive model is the seven-factor model of temperament
and character as assessed by the Temperament and
Character Inventory (TCI; Cloninger et al. 1993). The
TCI includes dimensions neglected in alternative fivefactor models (Svrakic et al. 1993; Zuckerman and
Cloninger 1996), particularly Self-Transcendence,
which is important in the development of well-being
(Cloninger 2004). The TCI distinguishes between temperament, defined as individual differences biasing the
associative conditioning of responses to simple emotional stimuli, and character, defined as individual differences in the supervisory cognitive processes that
modulate emotional conflicts. The four TCI temperament dimensions are Harm Avoidance (i.e., anxious vs.
risk taking), Novelty Seeking (i.e., impulsive vs. rigid),
Reward Dependence (i.e., sociable vs. aloof), and Per-

MEASUREMENT OF PERSONALITY AND


ITS DISORDERS
Measurement is the crux of scientific analysis, so nothing can be stated rigorously about the genetics of personality without initial consideration of fundamental
measurement issues that are often ignored in clinical research and practice. Traditionally, personality disorders have been described as a set of discrete categories,
as in DSM-IV-TR (American Psychiatric Association
2000) and ICD-10 (World Health Organization 1992),
but taxonomic analyses have long shown that such categorical descriptions are inadequate (Eysenck 1986). Efforts to demonstrate the discreteness of different psychiatric disorders have led to inconsistent results both
for psychoses and for milder anxiety and depressive
disorders (Cloninger et al. 1985; Kendell 1982). Even
when it is found that intermediate or combined syndromes are relatively rare, the separation of groups
(e.g., patients with schizophrenia versus patients with
bipolar disorder) has been incomplete (Cloninger et al.
1985; Sigvardsson et al. 1986). For personality disorders,
there is no evidence whatsoever of discrete boundaries
between categories or even clusters, and most patients
with any personality disorder satisfy criteria for two or
more putative categories (Cloninger 1987, 2002b, 2004;
Cloninger and Svrakic 2000; Cloninger et al. 1993).
Fortunately, individual differences in personality
can be measured well in terms of multiple quantitative
143

144

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

sistence (i.e., overachieving vs. underachieving). The


three TCI character dimensions are Self-Directedness
(i.e., purposeful vs. aimless), Cooperativeness (i.e.,
helpful vs. hostile), and Self-Transcendence (i.e., insightful vs. repressive). Each of these seven dimensions has unique genetic determinants that modulate
specific brain circuitry carrying on distinct information-processing tasks (Cloninger 2002a; Gillespie et al.
2003; Gusnard et al. 2003; Turner et al. 2003). Initially
temperament traits were expected to be more heritable
than character traits, but nothing about the model required this to be so. Empirical studies have now shown
that temperament and character differ in terms of their
psychological functions and brain circuitry but not in
their degree of heritability (Cloninger 2004).
Extensive clinical research has shown that putative
categories of personality disorder have a characteristic
multidimensional profile that allows diagnoses to be
made without the redundancy that results from lists of
categorical items (Cloninger 2000b). For example,
Cluster A personality disorders are distinguished by
low Reward Dependence (e.g., aloofness), Cluster B by
high Novelty Seeking (e.g., anger), and Cluster C by
high Harm Avoidance (e.g., anxiety). Empirically,
when a fourth cluster is identified, it is distinguished
by high Persistence as seen in anankastic or obsessional personality disorders. Individuals often satisfy
criteria for disorders in multiple clusters because the
four temperaments are nearly uncorrelated with one
another phenotypically and have little or no genetic
overlap. The clinical and genetic independence of the
temperament dimensions means that all possible combinations of high or low scores occur. For example, individuals with borderline or explosive temperament
configurations are defined as those individuals high in
both Harm Avoidance and Novelty Seeking and low in
Reward Dependence. Such explosive individuals have
strong approach-avoidance conflicts and emotional instability, which is characteristic of both Clusters B and
C in DSM-IV-TR (Cloninger 1987; Cloninger and
Svrakic 2000). Such multidimensional patterns of temperament and character influence susceptibility to specific patterns of comorbid psychopathology (Cloninger
2002b; Cloninger et al. 1998). Personality configurations are predictive of susceptibility to illness and patterns of comorbidity throughout the full range of psychopathology (Battaglia et al. 1996; Cloninger et al.
1994), including anxiety disorders (Jiang et al. 2003),
eating disorders (Anderson et al. 2002), substance dependence (Howard et al. 1997), somatoform disorders
(Cloninger 1986), mood disorders (Cloninger et al.
1998), and schizophrenia (Szoke et al. 2002).

John Livesley and his colleagues (1993) developed


the Dimensional Assessment of Personality Pathology
(DAPP) to measure the self-report of a hierarchy of dimensions of personality problems. The questionnaire
is composed of 560 items measuring 18 dimensions derived by factor analysis, each with at least three specific
facet scales. Factor analysis of the 18 basic scales yields
four higher-order factors labeled Emotional Lability or
Dysregulation, Antagonism or Dissocial Behavior, Interpersonal Unresponsiveness or Inhibition, and Compulsivity (Jang et al. 1996; Livesley et al. 1998). These
higher-order dimensions of abnormal personality resemble the dimensions of normal personality, such as
the dimensions of the five-factor model (FFM), which
includes Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness as assessed in the NEO personality inventory (Costa and
McCrae 1990). For example, DAPP Emotional Lability
or Dysregulation is similar to high Neuroticism in the
FFM and low Self-Directedness in the TCI. Antagonism or Dissocial Behavior is similar to low Agreeableness in the FFM and low Cooperativeness in the TCI.
Hence these dimensions define healthy personality at
one extreme (namely, high TCI Self-Directedness and
Cooperativeness, or low DAPP Emotional Dysregulation and Dissocial Behavior, or low NEO Neuroticism
and Agreeability) and personality disorder at the other
extreme (e.g., low TCI Self-Directedness and Cooperativeness) (Livesley et al. 1998; Svrakic et al. 1993).

CORRESPONDENCE BETWEEN NORMAL


AND ABNORMAL PERSONALITY STRUCTURE
One of the most robust findings about personality, but
one that is surprising to many psychiatrists, is that the
same dimensions of personality observed in the general population account well for the personality variation observed in psychiatric patients. Patients with
personality disorders and other forms of psychopathology have extreme values on one or more personality dimensions, but the dimensional structure is the
same in samples from the general community and
from psychiatric treatment facilities (Cloninger et al.
1994; Krueger 1996, 1999a, 1999b; Livesley et al. 1998;
Sigvardsson et al. 1987; Svrakic et al. 1993). This
shared structure is surprising to those who assume
that personality only colors the expression of mental
disorders, which are assumed to be independent and
discrete entities. Rather, many studies have demonstrated either strong correlations among measures of

Genetics

normal and abnormal personality (Cloninger and


Svrakic 1994; Costa and McCrae 1990; Duggan et al.
2003) or joint loadings of measures of normal and abnormal personality on the same factors (DiLalla et al.
1993). The genetic structure of normal and abnormal
personality traits in studies of twins is indistinguishable, suggesting that influences on normal and abnormal personality act through systems common to both,
whether the twins are reared together (Livesley et al.
1998) or apart (Markon et al. 2002). Furthermore, the
genetic and environmental structure of personality is
indistinguishable from that of common mental disorders (Krueger 1996, 1999b). In other words, the risk of
common mental disorders can be well explained as
clinical manifestations of personality traits. Childhood
personality is also moderately predictive of adult personality and psychopathology, suggesting that personality measures susceptibility to psychopathology
to the degree that personality and psychopathology
are heritable (Krueger 1996, 1999a; Sigvardsson et al.
1987). Together, these replicable findings show that
normal personality, abnormal personality, and common mental disorders share a common causal foundation. In other words, the same biopsychosocial
systems influence individual differences in normal
personality and its disorders.

THE INHERITANCE OF PERSONALITY


DISORDER SYMPTOMS AND CATEGORIES
Prior to 1993, most work on the inheritance of personality disorders was carried out with categorical diagnoses, as reviewed elsewhere (Thapar and McGuffin
1993). Family, twin, and adoption studies provided
clear evidence for the moderate heritability of schizotypal personality disorder and antisocial personality
disorder (ASPD), which were the most studied personality disorders up to that time. The most thoroughly studied personality disorder is ASPD, which is
commonly ascertained through studies of convicted
criminals. Adoption studies of ASPD (Crowe 1972,
1974), psychopathy (Schulsinger 1972), and criminality (Cloninger et al. 1975c; Mednick et al. 1984) were
carried out in the United States and Scandinavia, all
showing substantial heritable influences. Interactions
between genetic predisposition and childhood rearing
in an unstable environment were also demonstrated
for petty criminality and ASPD (Cadoret et al. 1985;
Cloninger et al. 1975c). Monozygotic (MZ) twins were
also more often concordant for criminality than were

145

dizygotic (DZ) twins, as reviewed in detail elsewhere


(Cloninger and Gottesman 1987). The concordances
for criminality and ASPD were predicted by differences in severity of liability by a function proportional
to the inverse of the prevalence in the general population, indicating that susceptibility was inherited to
an underlying quantitative variable like personality
(Cloninger et al. 1975a, 1975b). For example, ASPD is
less frequent in women than in men, so antisocial
women have a stronger genetic loading (measured by
more antisocial relatives) than do antisocial men.
Schizotypal personality disorder has also been extensively studied as a categorical diagnosis in family,
twin, and adoption studies of schizophrenia. There is
a consistently greater incidence of schizotypal individuals among the relatives of schizophrenic probands than among normal control subjects but less
consistent evidence of an excess of schizophrenic individuals among the relatives of schizotypal probands,
as reviewed elsewhere (Thapar and McGuffin 1993).
This asymmetry in results may be explained by the
lower severity of liability represented by schizotypal
probands or heterogeneity within schizotypal personality disorder. More recent work using quantitative
multidimensional measures of schizotypy indicates
that schizotypy is multidimensional and genetically
heterogeneous. For example, the positive and negative
components of schizotypy are both moderately heritable and genetically independent, although each may
contribute to cognitive disorganization (Linney et al.
2003). Only the eccentric, affect-constricted aspect of
schizotypal personality disorder may be within the
heritable spectrum of schizophrenia (Lyons et al. 1994;
Squires-Wheeler et al. 1989; Torgersen et al. 2000).
Since 1993, the inheritance of personality disorders
has been studied primarily through use of quantitative scales that measure individual differences in the
number of symptoms of personality disorder, because
quantitative scales contain more information than is
present when a categorical diagnosis is made. Quantitative measurement has also led to information about
the full range of symptoms observed in personality
disorders as defined in international classifications.
An excellent set of studies has been carried out by
John Livesley and his colleagues using the DAPP. This
questionnaire, briefly described earlier in the chapter,
was developed on the basis of linear factor analyses
that identified a stable structure underlying personality disorders in clinical and nonclinical subjects (Jang
et al. 1996, 1998; Livesley et al. 1993, 1998). The median
heritability estimates from these twin studies was
45%. Most dimensions of personality disorder symp-

146

Table 91.

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

A study of categorical personality disorders assessed by SCID-II interviews in 92 MZ twin pairs and
129 DZ twin pairs in Norway

Diagnoses
Any personality disorder
Any Cluster A
Any Cluster B
Any Cluster C

MZ correlation

DZ correlation

Heritability

0.58 0.10
0.37 0.14
0.60 0.11
0.61 0.09

0.36 0.10
0.09 0.11
0.31 0.12
0.23 0.11

0.44 0.20
0.37 0.25
0.59 0.23
0.59 0.20

Note. DZ=dizygotic; MZ=monozygotic; SCID-II=Structured Clinical Interview for DSM-IV Axis II Personality Disorders.
Source. Adapted from Torgersen et al. 2000

toms show moderate heritability (40%60%), as


shown in Table 91 (Jang et al. 1996). Moderate heritability was also observed for most traits even after regression on the four higher-order factors previously
described, suggesting that each basic scale measures
genetic variability not explained by the higher-order
dimensions (Livesley et al. 1998). Overall, the genetic
and environmental structure of personality disorder
symptoms was indistinguishable from that of normal
personality, suggesting a continuity between normal
and disordered personality (Livesley et al. 1993, 1998).
Other studies using other instruments and age groups
have obtained similar results indicating both the moderate heritability of personality disorder symptoms
and continuity between normal and disordered personality (Coolidge et al. 2001; Markon et al. 2002; Samuels et al. 2000).
One twin study based on a wide range of categorical diagnoses has been carried out in Norway by Svenn
Torgersen and his colleagues (Torgersen et al. 2000).
They identified 92 MZ twin pairs and 129 DZ twin
pairs in which at least one proband had a diagnosis of
personality disorder, and they divided the cases according to DSM clusters and categories. The probandwise concordances for any definite personality disorder were 40% for MZ pairs and 29% for DZ pairs, indicating substantial genetic influences (P<0.01), as summarized in Table 92. Concordance for membership in
personality disorder clusters also could not be explained without taking genetic variability into account.
Estimates of heritability for the clusters and specific
categories were moderate (i.e., between 40% and 60%),
much as observed for quantitative measures of normal
and abnormal personality traits (Eaves et al. 1989; Jang
et al. 1996). The correlations between DZ twins were
usually less than half of those of MZ pairs, suggesting
that genegene and geneenvironment interactions are
as important for categorical diagnoses as they are for
personality dimensions in twins reared apart (Bergeman et al. 1993; Pedersen et al. 1991; Plomin et al. 1998;

Tellegen et al. 1988). The small number of individuals


with particular categorical diagnoses made any conclusions about the heritability of individual categories
of personality in Torgersens twin study imprecise, as
shown by the large standard errors even for broad clusters of diagnoses shown in Table 92. For example,
most of the individuals with a Cluster B personality
disorder in Torgersens study had borderline personality disorder. As a result, the authors found significant
evidence that genetic factors influence vulnerability to
borderline personality disorder, but the estimate of the
magnitude of its heritability is imprecise due to the
small number of cases. Nevertheless, the consistency of
the overall results with those of studies of personality
models derived by linear factor analysis indicates that
personality and its disorders are moderately heritable
and approximately equally influenced by genetic factors and other factors unique to each individual. The
influence of environmental influences shared by siblings reared together accounts for less than 10% of the
variance in personality and its disorders (Eaves et al.
1989; Jang et al. 1996; Livesley et al. 1993, 1998; Pedersen et al. 1988).

THE INHERITANCE OF GENERAL


PERSONALITY DIMENSIONS
The personality dimensions relevant to the regulation
of gene expression in the brain are expected to be interactive with one another because of the nonlinear
nature of complex adaptive systems and extensive evidence that the extremes of personality dimensions are
reproductively disadvantageous (Cloninger 2000a).
The nonlinear nature of personality was foreseen in
the development of the seven-factor model of personality assessed by the TCI (Cloninger et al. 1993). The
TCI was developed as a set of scales measuring specific psychological constructs. No effort was made to

147

Genetics

Table 92.

Heritability and concordances in 236 MZ twin pairs and 247 DZ twin pairs for the 18 basic scales of
the Differential Assessment of Personality Pathology in Canada

Scale label
Affective lability
Anxiousness
Callousness
Cognitive distortion
Compulsivity
Conduct problems
Identity problems
Insecure attachment
Intimacy problems
Narcissism
Oppositionality
Rejection
Restricted expression
Self-harm
Social avoidance
Stimulus-seeking
Submissiveness
Suspiciousness

MZ correlation

DZ correlation

Heritability

0.49
0.42
0.56
0.48
0.40
0.53
0.51
0.45
0.47
0.51
0.41
0.33
0.48
0.39
0.52
0.38
0.41
0.42

0.12
0.25
0.32
0.31
0.19
0.36
0.28
0.27
0.24
0.22
0.29
0.19
0.26
0.26
0.27
0.21
0.29
0.29

0.45
0.44
0.56
0.49
0.37
0.56
0.53
0.48
0.48
0.53
0.46
0.35
0.41
0.41
0.53
0.40
0.45
0.45

Note. DZ=dizygotic; MZ=monozygotic.


Source. Adapted from Jang et al. 1996.

select or combine items in such a way as to give the appearance that the relations between the dimensions
were linear or functionally independent, as is done in
tests derived by factor analytic methods. Each of the
seven dimensions can be described as unique because
most of the correlation coefficients among dimensions
are negligible (< 0.25) and none is strong (>0.70). Nevertheless, each dimension has a correlation with at
least one other dimension, and these vary in magnitude from weak (0.250.39) to moderate (0.400.59).
Initial twin studies were carried out using only
measures of temperament (Heath et al. 1994; Stallings
et al. 1996). More recent studies have used the TCI and
show that each of the seven TCI dimensions has a
unique genetic variance that is not explained by the
other dimensions (Gillespie et al. 2003). The TCI includes heritable dimensions of personality that are neglected in five-factor models, including TCI Self-Transcendence and TCI Reward Dependence (Svrakic et al.
1993; Zuckerman and Cloninger 1996). A sample of
2,517 Australian twins aged 50 years or older between
1993 and 1995 completed the TCI. The correlation between each of the seven TCI dimensions of personality
was higher in MZ twin pairs than in DZ twin pairs,
suggesting significant genetic effects for each dimension. Heritability was derived using a standard multivariate model that is similar to estimating total genetic

effects from twice the difference between the correlations of MZ minus DZ twin pairs. For example, for
Harm Avoidance the correlations in female MZ twins
was 0.47 and that in female DZ twins was 0.21, so
twice the difference of 0.26 gives an estimate of 52%
heritability. The estimates of heritability for each TCI
dimension based on both male and female twins are
summarized in Table 93. Total genetic effects or heritability varied from 27%45%, without correcting for
the reliability of the short forms of the TCI used in this
study, and correction for reliability showed that each
dimension was moderately heritable. Each dimension
still had significant unique genetic variance when any
overlap with other dimensions was taken into account. Both additive genetic and environmental influences unique to each individual were significant. Environmental influences shared by twin pairs reared
together did not improve the fit of the model to the
data. Hence, higher cognitive processing, as measured
by TCI character traits, was found to be as heritable as
emotional processing, as measured by TCI temperament traits. Nongenetic influences unique to each individual influence each dimension of personality as
much as do genetic factors.
These studies of twins did suggest substantial genetic effects on human personality, warranting subsequent molecular genetic studies of linkage and associ-

148

Table 93.

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Total genetic effects (heritability) of each


of the seven TCI personality dimensions
estimated in 2,517 twins in Australia

Personality dimension

Harm Avoidance
Novelty Seeking
Reward Dependence
Persistence
Self-Directedness
Cooperativeness
Self-Transcendence

Genetic effects
Total
percentage

Unique
percentage

42
39
35
30
34
27
45

29
32
20
23
25
16
26

Note. Unique effects exclude genetic contributions shared with


other personality dimensions.
TCI= Temperament and Character Inventory.
Source. Adapted from Gillespie et al. 2003.

ation (Benjamin et al. 2002; Cloninger 1998). Many


twin studies suggest heritabilities of about 50% for
most complex personality and cognitive traits, including the dimensions of the three-, five-, and sevenfactor models of personality (Eaves et al. 1989; Gillespie et al. 2003; Heath et al. 1994; Pedersen et al. 1988;
Stallings et al. 1996). However, these estimates of additive genetic effects are inflated by genegene and
geneenvironmental interactions that have a greater
influence on MZ than on DZ twins (Cloninger et al.
1979). These interactions indicate that particular combinations of genetic and environmental factors interact in ways that cannot be explained by their average
effectsthat is, the whole is more than the sum of its
parts. Adoption studies indicate that the heritability of
personality is about 20%30% (Cloninger 1998; Loehlin 1992; Plomin et al. 1998) rather than 50%. The discrepancy between twin and adoption studies suggests
that the estimates of heritability in twin studies are inflated by genegene and geneenvironment interactions or that the estimates in adoption studies are reduced by the effects that distinguish the members of
two generations, such as age and cohort effects. Either
explanation involves nonadditive interactions among
multiple genetic and environmental factors; thus, the
whole personality is more than the sum of its individual parts. Unfortunately, twin studies have little or no
power to test their Mendelian assumptions, such as
the assumption that the total genetic effects are additive (Gillespie et al. 2003; Plomin et al. 1998; Torgersen
et al. 2000). A practical consequence of these limitations is that the probable nonadditive interactions between genes must be examined directly by measuring

specific genetic polymorphisms in molecular genetic


studies of linkage and association (Benjamin et al.
2002) and then carrying out studies of geneenvironment interaction (Caspi et al. 2003; Keltikangas-Jarvinen et al. 2004). There is now substantial direct evidence that personality development depends on the
nonadditive effects of genegene (Benjamin et al. 2000;
Strobel et al. 2003; Van Gestel et al. 2002) and gene
environment (Caspi et al. 2003; Keltikangas-Jarvinen
et al. 2004) interactions.
The interactions among a few genes that have been
extensively studied in relation to personality illustrate
the complex interactions among variables influencing
the development of personality and its relationship to
psychopathology. Serotonin and dopamine are phylogenetically ancient neurotransmitters that play basic
roles in brain function and behavior (Cravchik and
Goldman 2000). Extensive diversity in the regulation
of serotonin and dopamine function has been studied
sufficiently in relation to personality to identify some
general epigenetic principles. Some of the genetic
polymorphisms that have been studied most extensively are genes that promote, transport, and catabolize these key neurotransmitters. The marked differences in such neurotransmitter functions between
individuals is of special interest because they modulate individual differences in personality (Borg et al.
2003; Cloninger 1987; Ding et al. 2002; Hamer et al.
1999; Strobel et al. 2003).
In the human prefrontal cortex, the enzyme catechol-O-methyltransferase (COMT) is critical in the
metabolic degradation of dopamine, and by regulating dopamine availability, it can influence personality
and cognitive function. The COMT gene has a common variant (i.e., polymorphism) involving the substitution of valine (Val) by methionine (Met) at position
158. Individuals who are homozygous Met/Met have
fourfold less activity than Val/Val or Val/Met individuals. Individuals with the low-activity form of COMT
(i.e., Met/Met homozygotes) have been found to have
higher scores in TCI Harm Avoidance and to have
low-voltage alpha activity on electroencephalography
more frequently than others (Enoch et al. 2003).
The expression of the serotonin transporter is regulated by a promoter polymorphism that is unique in humans and simian primates (Reif and Lesch 2003). Its
low-activity form (i.e., short allele) interacts nonlinearly
with stressful life events to increase susceptibility to depression, illustrating the importance of geneenvironment interaction (Caspi et al. 2003). The short allele has
been associated with anxiety-related traits such as neuroticism, which confound high Harm Avoidance with

Genetics

the low Self-Directedness and low Cooperativeness


seen in personality disorders. The less active short allele
of the serotonin transporter promoter may be more
strongly associated with low TCI Self-Directedness and
low TCI Cooperativeness than with the temperament
trait of Harm Avoidance (Hamer et al. 1999; Thierry et
al. 2004).
Furthermore, the high-activity forms of COMT
(Val/Val or Val/Met) have increased dopamine catabolism in the prefrontal cortex, which impairs prefrontal
brain physiology and increases perseverative errors in
the Wisconsin Card Sorting Test, and which has been
found to slightly increase the risk of schizophrenia
(Egan et al. 2001). The genotype explained 4% of the
variance in perseverative errorsthat is, having difficulty switching strategies when an error is made. The
relationship was further supported by studies of prefrontal physiology on functional magnetic resonance
imaging during a working memory task and by a family study showing the increased transmission of the
Val allele to schizophrenic offspring (Egan et al. 2001).
Such findings about a single gene have often
proven to be inconsistently replicable when they are
studied by many independent groups, as has been the
case with the association of the dopamine receptor
gene DRD4 and Novelty Seeking (Kluger et al. 2002;
Schinka et al. 2002). Such inconsistent results need to
be evaluated within the general context of the role of
personality dimensions as moderator variables in
nonlinear adaptive systems. As a result of their nonlinear function as moderators, the inheritance of personality is expected to involve genegene and gene
environment interactions (Keltikangas-Jarvinen et al.
2004; Kluger et al. 2002). For most quantitative traits,
individuals with intermediate values are usually
adapted better than individuals with extremely high
or low values. In contrast, individuals at the each extreme of a quantitative trait are more prone to disorders and are less well adapted than intermediate individuals. For example, type 1 and type 2 alcoholism
may be at opposite extremes of the epigenetic regulation of alcohol consumption, with type 1 alcoholics being mature but excessively Harm Avoidant and type 2
alcoholics being immature and excessively Novelty
Seeking (Cloninger 2004). Other intermediate personality profiles provide sufficient checks and balances to
maintain normative drinking, but both extremes increase susceptibility to problems. Unless the relevant
interacting biopsychosocial variables are simultaneously measured, results with individual variables in
different samples are expected to be inconsistent despite their validity in some contexts.

149

Molecular genetic studies on Novelty Seeking confirm the importance of nonlinear genegene and
geneenvironment interactions in Novelty Seeking. A
polymorphism of the dopamine transporter is associated with individual differences in initiating and continuing to smoke cigarettes, effects which are mediated by the joint association of cigarette smoking and
the dopamine transporter with Novelty Seeking
(Sabol et al. 1999). In addition, the DRD4 exon 2 sevenrepeat allele has been associated with high Novelty
Seeking and increased risk of opiate dependence (Kotler et al. 1997). Other work has shown that Novelty
Seeking is associated with the ten-repeat allele of the
dopamine transporter DAT1 when the DRD4 sevenrepeat allele is absent (Van Gestel et al. 2002).
Novelty Seeking also depends on the three-way
interaction of DRD4 with COMT and the serotonin
transporter locus promoter s regulatory region
(5-HTTLPR). In the absence of the short 5-HTTLPR allele (5-HTTLPR L/L genotype) and in the presence of
the high-activity COMT Val/Val genotype, Novelty
Seeking scores are higher in the presence of the DRD4
seven-repeat allele than its absence (Benjamin et al.
2000). Furthermore, within families, siblings who
shared identical genotype groups for all three polymorphisms (COMT, DRD4, and 5-HTTLPR) had significantly correlated Novelty Seeking scores (intraclass correlation = 0.39 in 49 subjects, P < 0.008). In
contrast, siblings with dissimilar genotypes in at least
one polymorphism showed no significant correlation
for Novelty Seeking (intraclass coefficient= 0.18 in 110
subjects, P = 0.09). Similar interactions were also observed between these three polymorphisms and Novelty Seeking in an independent sample of unrelated
subjects (Benjamin et al. 2000) and have been replicated by independent investigators (Strobel et al.
2003). A similar three-way interaction has been described for the temperament dimension Persistence
with D4DR and D3DR and the serotonin receptor gene
type 2c (5-HT2c) (Ebstein et al. 1997).
Geneenvironment interaction has also been demonstrated for TCI Novelty Seeking in prospective population-based studies (Ekelund et al. 1999; KeltikangasJarvinen et al. 2003, 2004). The TCI was administered to
two large birth cohorts of Finnish men and women, and
the individuals who scored in the top 10% and bottom
10% of TCI Novelty Seeking were genotyped for the
exon 3 repeat polymorphism of DRD4. The four-repeat
and seven-repeat alleles were most common in the
Finnish sample (Ekelund et al. 1999; Keltikangas-Jarvinen et al. 2003), as is usual throughout the world
(Ding et al. 2002). The two-repeat and five-repeat al-

150

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

leles, which are rare in the Americas and Africa, were


more than three times as frequent (16% vs. 5%) in Finns
who were very high in Novelty Seeking than in those
who were very low in Novelty Seeking (Ekelund et al.
1999; Keltikangas-Jarvinen et al. 2003), and this difference was replicated in an independent sample (Keltikangas-Jarvinen et al. 2003). The association with the
two-repeat and five-repeat alleles was strongest for the
two most adaptive aspects of Novelty Seeking, exploratory excitability and impulsive decision making (Keltikangas-Jarvinen et al. 2003). Finnish men and women
with the two-repeat and five-repeat alleles of the exon 3
DRD4 polymorphism were higher in Novelty Seeking
as adults if they experienced a hostile childhood environment, as measured by maternal reports of emotional
distance and a strict authoritarian disciplinary style
with physical punishment (Keltikangas-Jarvinen et al.
2004). The mothers reports of childhood environment
were obtained when the children were aged 1821
years, and genotyping and personality assessment of
Novelty Seeking was done independently 15 years
later. If children had the two-repeat or five-repeat alleles of the DRD4 polymorphism, their TCI Novelty
Seeking scores were high if they were reared in a hostile
childhood environment and their Novelty Seeking was
low if they were reared in a kind and cooperative environment. Children with certain genotypes are likely to
evoke a characteristic pattern of responses from their
parents and others, and to select for themselves certain
aspects from the available environments (Scarr and McCartney 1983). However, therapeutic environments,
such as kind and cooperative parenting, can evoke positive adaptation by modifying gene expression, which
depends on the orchestrated interaction of many genes
and environmental influences (Keltikangas-Jarvinen et
al. 2004). Such complex genegene and geneenvironment interactions are well documented for other common diseases, such as coronary artery disease and hypertension (Sing et al. 1996; Zerba et al. 2000). For
example, a hostile childhood environment is correlated
with many variables that are associated with coronary
artery disease, such as marital dissatisfaction, type A
personality, emotional distance, and a high level of job
involvement among fathers (Keltikangas-Jarvinen et al.
2004).
What are the consequences of such complex epigenetic effects for brain activity? Fortunately, the relationship between individual differences in Persistence and
the activity of related brain networks has been worked
out in detail recently (Gusnard et al. 2003). Persistence
was found to be strongly correlated (r=0.79) with activation of a well-known circuit for regulation of reward-

seeking behavior in a recent functional magnetic resonance imaging study (Gusnard et al. 2003). The circuit
includes the ventral striatum, anterior cingulate (Brodmann area 24), and orbitofrontal cortex (Brodmann area
47) bilaterally. Subjects were asked to rate pictures as
pleasant, unpleasant, or neutral. As the percentage of
neutral pictures in the picture set increased, subjects
who were more persistent rated more pictures as pleasant at the expense of neutral pictures (r=0.34, P<0.05).
This selection bias was independent of the percentage
of neutral pictures in the sets. There was also a nonlinear interaction between the percentage in brain activity
change and the percentage of neutral pictures. The
same distributed neural circuit was upregulated (i.e.,
became more metabolically active) when persistent individuals viewed a high percentage of neutral pictures,
but it was downregulated in nonpersistent individuals
when they viewed a high percentage of neutral pictures.
These findings show that the regulation of gene expression by personality is mediated by complex adaptive systems made up of multiple genetic and environmental factors. Personality is made up of multiple
heritable dimensions composed of unique but partially overlapping sets of epistatic genes (i.e., genes
that interact nonadditively with other genes). These
developmental systems modulate brain states by
modifying the transitory connections between changing distributed networks of neurons. The prominence
of genegene and geneenvironment interactions is
characteristic of most common diseases and quantitative phenotypes such as personality traits (Cloninger
2004).

IMPLICATIONS FOR DIAGNOSIS AND


CLINICAL PRACTICE
A substantial body of research now shows that personality and its disorders are complex biopsychosocial phenomena influenced by multiple genetic and
environmental variables. The psychobiological systems that regulate personality are under stabilizing
selection in which intermediate phenotypes are favored under most circumstances. Consequently, individuals in the extremes of these complex adaptive networks are vulnerable to psychopathology. To be more
specific, the TCI measures seven major dimensions of
personality that account for individual differences in
both normal and disordered personality traits. Each of
these dimensions of personality is influenced by dif-

Genetics

ferent genes and by variables unique to each individual. Extreme configurations of these dimensions are
vulnerable to personality disorders and related psychopathology. Consequently, clinicians can efficiently
assess personality and susceptibility to other psychopathology by measurement of variables that can be assessed in every patient as part of an adequate mental
status examination and psychiatric history (Cloninger
2000b).
Temperament is moderately stable regardless of
treatment but is modulated by character traits that determine the level of a persons maturity and integration. Antidepressants and cognitive behavioral therapy have been shown to increase a persons SelfDirectedness (Anderson et al. 2002; Cloninger 2000b;
Tome et al. 1997). Such increases in self-awareness initiate a self-organizing spiral of development leading
to increasing maturity and well-being (Cloninger
2004). A unified biopsychosocial approach to personality and its disorders is possible by understanding
the normal path of personality development along
with the genetic and environmental influences that
may lead to deviations from that path. The study of
the psychobiology of human personality provides the
foundation for any coherent understanding of mental
disease and mental health.

REFERENCES
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association,
2000
Anderson CB, Joyce PR, Carter FA, et al: The effect of cognitive-behavioral therapy for bulimia nervosa on temperament and character as measured by the temperament and
character inventory. Compr Psychiatry 43:182188, 2002
Battaglia M, Przybeck TR, Bellodi L, et al: Temperament dimensions explain the comorbidity of psychiatric disorders. Compr Psychiatry 37:292298, 1996
Benjamin J, Osher Y, Kotler M, et al: Association of tridimensional personality questionnaire (TPQ) traits and
three functional polymorphisms: dopamine receptor
D4 (DRD4), serotonin transporter promoter region
(5-HTT LPR) and catecho l O -met hylt ransferase
(COMT). Mol Psychiatry 5:96100, 2000
Benjamin J, Ebstein RP, Belmaker RH (eds): Molecular Genetics and the Human Personality. Washington, DC,
American Psychiatric Publishing, 2002
Bergeman CS, Chipur HM, Plomin R, et al: Genetic and environmental effects on openness to experience, agreeableness, and conscientiousness: an adoption/twin
study. J Pers 61:159179, 1993

151

Borg J, Andree B, Soderstrom H, et al: The serotonin system


and spiritual experiences. Am J Psychiatry 160:1965
1969, 2003
Cadoret RJ, OGorman TW, Troughton E, et al: Alcoholism
and antisocial personality: interrelationships, genetics
and environmental factors. Arch Gen Psychiatry 42:162
167, 1985
Caspi A, Sugden K, Moffitt TE, et al: Influence of life stress
on depression: moderation by a polymorphism in the
5-HTT gene. Science 301:386389, 2003
Cloninger CR: A unified biosocial theory of personality and
its role in the development of anxiety states. Psychiatr
Dev 3:167226, 1986
Cloninger CR: A systematic method for clinical description
and classification of personality variants: a proposal.
Arch Gen Psychiatry 44:573587, 1987
Cloninger CR: The genetics and psychobiology of the seven
factor model of personality, in The Biology of Personality Disorders. Edited by Silk KR. Washington, DC,
American Psychiatric Press, 1998, pp 6384
Cloninger CR: Biology of personality dimensions. Curr Opin
Psychiatry 13:611616, 2000a
Cloninger CR: A practical way to diagnose personality disorder: a proposal. J Personal Disord 14:99108, 2000b
Cloninger CR: Functional neuroanatomy and brain imaging
of personality and its disorders, in Biological Psychiatry, Vol 2. Edited by DHaenen H, den Boer JA, Willner
P. Chichester, England, Wiley, 2002a, pp 13771385
Cloninger CR: Implications of comorbidity for the classification of mental disorders: the need for a psychobiology
of coherence, in Psychiatric Diagnosis and Classification. Edited by Maj M, Gaebel W, Lopez-Ibor JJ, et al.
Chichester, England, Wiley, 2002b, pp 79106
Cloninger CR: Feeling Good: The Science of Well Being.
New York, Oxford University Press, 2004
Cloninger CR, Gottesman II: Genetic and environmental factors in antisocial behavior disorders, in Causes of
Crime: New Biological Approaches. Edited by Mednick
SA, Moffitt TE, Stack SA. Cambridge, England, Cambridge University Press, 1987, pp 92109
Cloninger CR, Svrakic DM: Differentiating normal and abnormal personality by the seven-factor personality model, in
Differentiating Normal and Abnormal Personality. Edited by Strack S, Lorr M. New York, Springer, 1994
Cloninger CR, Svrakic DM: Personality disorders, in Comprehensive Textbook of Psychiatry. Edited by Sadock BJ, Sadock VA. New York, Lippincott Williams and Wilkins,
2000, pp 17231764
Cloninger CR, Reich T, Guze SB: The multifactorial model of
disease transmission, II: sex differences in the familial
transmission of sociopathy (antisocial personality). Br
J Psychiatry 127:1122, 1975a
Cloninger CR, Reich T, Guze SB: The multifactorial model of
disease transmission, III: familial relationship between
sociopathy and hysteria (Briquets syndrome). Br J Psychiatry 127:2332, 1975b

152

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Cloninger CR, Sigvardsson S, Bohman M, et al: Predisposition to petty criminality in Swedish adoptees, II: crossfostering analysis of gene-environment interaction.
Arch Gen Psychiatry 39:12421247, 1975c
Cloninger CR, Rice J, Reich T: Multifactorial inheritance with
cultural transmission and assortative mating, II: a general model of combined polygenic and cultural inheritance. Am J Hum Genet 31:176198, 1979
Cloninger CR, Martin RL, Guze SB, et al: Diagnosis and
prognosis in schizophrenia. Arch Gen Psychiatry 42:12
25, 1985
Cloninger CR, Svrakic DM, Przybeck TR: A psychobiological model of temperament and character. Arch Gen Psychiatry 50:975990, 1993
Cloninger CR, Przybeck TR, Svrakic DM, et al: The Temperament and Character Inventory: A Guide to Its Development and Use. St. Louis, MO, Washington University
Center for Psychobiology of Personality, 1994
Cloninger CR, Bayon C, Svrakic DM: Measurement of temperament and character in mood disorders: a model of
fundamental states as personality types. J Affect Disord
51:2132, 1998
Coolidge FL, Thede LL, Jang KL: Heritability of personality
disorders in childhood: a preliminary investigation.
J Personal Disord 15:3340, 2001
Costa PT, McCrae RR: Personality disorders and the fivefactor model of personality. J Personal Disord 4:362371,
1990
Cravchik A, Goldman D: Neurochemical individuality: genetic diversity among human dopamine and serotonin
receptors and transporters. Arch Gen Psychiatry 57:1105
1114, 2000
Crowe RR: The adopted offspring of women criminal offenders: a study of their arrest records. Arch Gen Psychiatry 27:600603, 1972
Crowe RR: An adoption study of antisocial personality. Arch
Gen Psychiatry 31:785791, 1974
DiLalla DL, Gottesman II, Carey G: Assessment of normal
personality in an abnormal sample, in Progress in Experimental Personality and Psychopathology Research,
Vol 16. Edited by Chapman LJ, Chapman JP, Fowles D.
New York, Springer, 1993, pp 137162
Ding YC, Chi HC, Grady DL, et al: Evidence for positive selection acting at the human dopamine receptor D4 gene
locus. Proc Natl Acad Sci U S A 99:309314, 2002
Duggan C, Milton J, Egan V, et al: Theories of general personality and mental disorder. Br J Psychiatry 182 (suppl
44):S19S23, 2003
Eaves LJ, Eysenck HJ, Martin NG: Genes, Culture and Personality: An Empirical Approach. London, England,
Academic Press, 1989
Ebstein RP, Segman R, Benjamin J, et al: 5-HT 2C serotonin receptor gene polymorphism associated with the human
personality trait of reward dependence: interaction
with dopamine D4 receptor (D4DR) and dopamine D3
receptor (D3DR) polymorphisms. Am J Med Genet 74:
6572, 1997

Egan MF, Goldberg TE, Kolachana BS, et al: Effect of COMT


Val108/158Met genotype on frontal lobe function and
risk for schizophrenia. Proc Natl Acad Sci U S A 98:6917
6922, 2001
Ekelund J, Lichtermann D, Jaervelin M-R, et al: Association
between novelty seeking and the type 4 dopamine receptor gene in a large Finnish cohort sample. Am J Psychiatry 156:14531455, 1999
Enoch MA, Xu K, Ferro E, et al: Genetic origins of anxiety in
women: a role for a functional catechol-O-methyltransferase polymorphism. Psychiatr Genet 13:3341, 2003
Eysenck HJ: A critique of contemporary classification and
diagnosis, in Contemporary Directions in Psychopathology. Edited by Millon T, Klerman G. New York,
Guilford, 1986, pp 7398
Gillespie NA, Cloninger CR, Heath AC, et al: The genetic
and environmental relationship between Cloningers
dimensions of temperament and character. Pers Individ
Dif 35:19311946, 2003
Gusnard DA, Ollinger JM, Shulman GL, et al: Persistence
and brain circuitry. Proc Natl Acad Sci U S A 100:3479
3484, 2003
Hamer DH, Greenberg BD, Sabol SZ, et al: Role of the serotonin transporter gene in temperament and character.
J Personal Disord 13:312327, 1999
Heath AC, Cloninger CR, Martin NG: Testing a model for
the genetic structure of personality: a comparison of the
personality systems of Cloninger and Eysenck. J Pers
Soc Psychol 66:762775, 1994
Howard MO, Kivlahan D, Walker RD: Cloningers tridimensional theory of personality and psychopathology: applications to substance use disorders. J Stud Alcohol
58:4866, 1997
Jang KL, Livesley WJ, Vernon PA, et al: Heritability of personality disorder traits: a twin study. Acta Psychiatr
Scand 94:438444, 1996
Jang KL, Livesley WJ, Vernon PA: A twin study of genetic
and environmental contributions to gender differences
in traits delineating personality disorder. Eur J Pers
12:331344, 1998
Jiang N, Sato T, Hara T, et al: Correlations between trait anxiety, personality and fatigue: study based on the Temperament and Character Inventory. J Psychosom Res
55:493500, 2003
Keltikangas-Jarvinen L, Elovainio M, Kivimaeki M, et al: Association between the type 4 dopamine receptor gene
polymorphism and novelty seeking. Psychosom Med
65:471476, 2003
Keltikangas-Jarvinen L, Raeikkoenen K, Ekelund J, et al: Nature and nurture in novelty seeking. Mol Psychiatry
9:308311, 2004
Kendell RE: The choice of diagnostic criteria for biological
research. Arch Gen Psychiatry 39:13341339, 1982
Kluger AN, Siegfried Z, Ebstein RP: A meta-analysis of the
association between DRD4 polymorphism and novelty
seeking. Mol Psychiatry 7:712717, 2002

Genetics

Kotler M, Cohen H, Segman R: Excess dopamine D4 receptor (D4DR) exon III seven repeat allele in opioid-dependent subjects. Mol Psychiatry 2:251254, 1997
Krueger RF: Personality traits are differentially linked to
mental disorders: a multitrait-multidiagnosis study of
an adolescent birth cohort. J Abnorm Psychol 105:299
312, 1996
Krueger RF: Personality traits in late adolescence predict
mental disorders in early adulthood: a prospective epidemiological study. J Pers 67:3965, 1999a
Krueger RF: The structure of common mental disorders.
Arch Gen Psychiatry 56:921926, 1999b
Linney YM, Murray RM, Peters ER, et al: A quantitative genetic analysis of schizotypal personality traits. Psychol
Med 33:803816, 2003
Livesley WJ, Jang KL, Jackson DN, et al: Genetic and environmental contributions to dimensions of personality
disorder. Am J Psychiatry 150:18261831, 1993
Livesley WJ, Jang KL, Vernon PA: Phenotypic and genetic
structure of traits delineating personality disorder. Arch
Gen Psychiatry 55:941948, 1998
Loehlin JC: Genes and Environment in Personality Development. Newbury Park, CA, Sage, 1992
Lyons MJ, Toomey R, Faraone SV, et al: Comparison of
schizotypal relatives of schizophrenia versus affective
probands. Am J Med Genet 54:279285, 1994
Markon KE, Krueger RF, Bouchard TJJ, et al: Normal and abnormal personality traits: evidence for genetic and environmental relationships in the Minnesota Study of Twins
Reared Apart. J Pers 70:661693, 2002
Mednick SA, Gabrielli WFJ, Hutchings B: Genetic influences
in criminal convictions: evidence from an adoption cohort. Science 22:891894, 1984
Pedersen NL, Plomin R, McClearn GE, et al: Neuroticism,
extraversion, and related traits in adult twins reared
apart and reared together. J Pers Soc Psychol 55:950
957, 1988
Pedersen NL, McClearn GE, Plomin R, et al: The Swedish
Adoption Twin Study of Aging: an update. Acta Genet
Med Gemellol (Roma) 40:720, 1991
Plomin R, Corley R, Caspi A, et al: Adoption results for selfreported personality: evidence for nonadditive genetic
effects? J Pers Soc Psychol 75:211218, 1998
Reif A, Lesch KP: Toward a molecular architecture of personality. Behav Brain Res 139:120, 2003
Sabol SZ, Nelson ML, Fisher C, et al: A genetic association
for cigarette smoking behavior. Health Psychol 18:713,
1999
Samuels J, Nestadt G, Bienvenu OJ, et al: Personality disorders and normal personality dimensions in obsessivecompulsive disorder. Br J Psychiatry 177:457462, 2000
Scarr S, McCartney K: How people make their own environments: a theory of genotype greater than environment
effects. Child Dev 54:424435, 1983
Schinka JA, Letsch EA, Crawford FC: DRD4 and novelty
seeking: results of meta-analyses. Am J Med Genet
114:643648, 2002

153

Schulsinger F: Psychopathy, heredity and environment. Int


J Ment Health 1:190206, 1972
Sigvardsson S, Bohman M, von Knorring AL, et al: Symptom
patterns and causes of somatization in men. Genet Epidemiol 3:153169, 1986
Sigvardsson S, Bohman M, Cloninger CR: Structure and stability of childhood personality: prediction of later social
adjustment. J Child Psychol Psychiatry 28:929946, 1987
Sing CF, Haviland MB, Reilly SL: Genetic architecture of
common multifactorial diseases. Ciba Foundation Symposium 197:211229, 1996
Squires-Wheeler E, Skodal A, Bassett A, et al: DSM-III-R
schizotypal personality traits in offspring of schizophrenic disorder, affective disorder, and normal control
parents. J Psychiatr Res 23:229239, 1989
Stallings MC, Hewitt JK, Cloninger CR, et al: Genetic and environmental structure of the Tridimensional Personality
Questionnaire: three or four temperament dimensions?
J Pers Soc Psychol 70:127140, 1996
Strobel A, Lesch KP, Jatzke S, et al: Further evidence for a
modulation of Novelty Seeking by DRD4 exon III,
5-HTTLPR, and COMT val/met variants. Mol Psychiatry 8:371372, 2003
Svrakic DM, Whitehead C, Przybeck TR, et al: Differential
diagnosis of personality disorders by the seven factor
model of temperament and character. Arch Gen Psychiatry 50:991999, 1993
Szoke A, Schurhoff F, Ferhadian N, et al: Temperament in
schizophrenia: a study of the tridimensional personality
questionnaire. Eur Psychiatry 17:379383, 2002
Tellegen A, Lykken TD, Bouchard TJJ, et al: Personality similarity in twins reared apart and together. J Pers Soc Psychol 54:10311039, 1988
Thapar A, McGuffin P: Is personality disorder inherited? An
overview of the evidence. J Psychopathol Behav Assess
15:325345, 1993
Thierry N, Willeit M, Praschak-Rieder N, et al: Serotonin
transp orter pro moter gene po lymo rp hic region
(5-HTTLPR) and personality in female patients with
seasonal affective disorder and in healthy controls. Eur
Neuropsychopharmacol 14:5358, 2004
Tome MB, Cloninger CR, Watson JP, et al: Serotonergic autoreceptor blockade in the reduction of antidepressant
latency: personality and response to paroxetine and
pindolol. J Affect Disord 44:101109, 1997
Torgersen S, Lygren S, Oien PA, et al: A twin study of personality disorders. Compr Psychiatry 41:416425, 2000
Turner RM, Hudson IL, Butler PH, et al: Brain function and
personality in normal males: a SPECT study using statistical parametric mapping. Neuroimage 19:11451162,
2003
Van Gestel S, Forsgren T, Claes S, et al: Epistatic effects of
genes from the dopamine and serotonin systems on the
temperament traits of novelty seeking and harm avoidance. Mol Psychiatry 7:448450, 2002

154

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

World Health Organization: International Statistical Classification of Diseases and Related Health Problems, 1989
Revision, Geneva, Switzerland, World Health Organization, 1992

Zerba KE, Ferrell RE, Sing CF: Complex adaptive systems


and human health: the influence of common genotypes
of the apolipoprotein E (ApoE) gene polymorphism and
age on the relational order within a field of lipid metabolism traits. Hum Genet 107:466476, 2000
Zuckerman M, Cloninger CR: Relationships between Cloningers, Zuckermans, and Eysencks dimensions of personality. Pers Individ Dif 21:283285, 1996

10
Neurobiology
Emil F. Coccaro, M.D.
Larry J. Siever, M.D.

The study of personality disorder involves the study of


both disordered character and disordered temperament. Character relates to how we see and operate in our
world and is based on how we develop and what we
are taught about how to go through life. Temperament, in
contrast, relates to our innate tendency to behave and to
react to any of a variety of challenges presented by other
people and our environment. Although both aspects of
personality may be studied empirically, the study of
temperament is uniquely suited to biological study because temperament has known genetic and neurobiological correlates, both of which are linked to critical
processes involving cognition, emotion, and behavior.
The neurobiology of temperament, as it appears in
personality disorders, can be studied in a variety of
ways, including those that involve behavioral genetics,
neuropsychopharmacology and molecular genetics, and
psychophysiology and neuroimaging. Behavioral genetic study informs us about the degree to which personality (or temperamental) traits are under genetic influence. This work largely involves studies of families and
twins and is designed to document familial, if not genetic, components to behavior. Previous work defining
the genetic underpinnings of temperament has been critical to our current understanding that temperament is in-

herently biological in nature. Neuropsychopharmacological study informs us about the nature of brain
chemistry and how the regulation of any of a variety of
brain neurotransmitters influences temperament. Work
in this area has led to the understanding that brain serotonin, for example, is critical in modulating impulsive
aggressive behavior in individuals with personality disorder. Consequently, work in neuropsychopharmacology leads to work in molecular genetics whereby the
presence of a specific copy of a specific gene (e.g., for a
component of the brain serotonin system) influences a
temperamental trait. For example, individuals carrying a
specific gene for the serotonin transporter may be more
anxious than other individuals who do not carry this
gene. Finally, work in psychophysiology and neuroimaging brings investigative work up to a level that integrates genes, neuropsychopharmacology, and networks
of neural transmission. In this methodology, both brain
structure and brain function are examined regarding
their contribution to the expression of various temperamental traits. In some groups of patients with personality disorder, neuroimaging has revealed differences in
the size and function of specific structures.
Ultimately, the study of the neurobiology of personality disorders is conducted to lead to a more com155

156

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

prehensive understanding of the biological substrates


of personality disorder so that better treatments may
be discovered and/or so that existing treatments may
be improved. Uncovering the biological substrate for
a specific temperamental trait naturally leads to treatment strategies aimed at this specific substrate. The
best example of this approach is the use of serotonin
uptake inhibitors in the treatment of impulsive aggression in personality-disordered individuals. Curiously, work in this area revealed the likely presence of
two treatment response groups: one responsive to serotonin uptake inhibitors, the other responsive to
mood stabilizers.
In this chapter, we discuss the various aspects of the
neurobiology of personality disorder on a cluster-bycluster basis. We have chosen this organization because
the prototypical personality disorders of interest tend
to break out into one of the three personality disorder
clusters. Despite this type of organization, we should
note that research has clearly shown the relevance of a
dimensional approach to the study of personality. Each
section begins with a brief summary of the phenomenology characteristic of each personality disorder cluster and follows with a summary of data relevant to behavioral genetics, neuropsychopharmacology (and
molecular genetics where relevant), and neuropsychology and neuroimaging. Each of the first two sections
ends with a brief vignette illustrating some of the
points made about the psychobiology of prototypical
patients with selected personality disorders.

CLUSTER A PERSONALITY DISORDERS


The Cluster A personality disorders include schizotypal, paranoid, and schizoid personality disorder. The
criteria of these disorders capture shared characteristics
of social isolation, detachment, suspiciousness, and in
the case of schizotypal personality disorder, psychoticlike cognitive/perceptional distortion. Schizotypal personality was formulated in part on the clinical profile
observed in relatives of schizophrenic probands,
whereas the other two were defined more in a clinical
tradition. A high degree of overlap exists between
schizotypal and paranoid personality disorder, whereas
schizoid personality disorder is not frequently diagnosed in the clinical setting and may represent a milder
version of the Cluster A personality disorders. These
disorders can be perceived as consisting of a dimension
of social deficits (no friends, detached affect) and cognitive impairment, and in the case of schizotypal personality disorder, a psychotic-like dimension. Because of its

relationship to schizophrenia and its more common


prevalence in clinical populations, most of the neurobiological research on this cluster has focused on schizotypal personality disorder and is summarized here in relation to these dimensions.

Behavioral Genetics
Schizotypal personality disorder is found more frequently in the relatives of schizophrenic probands than
in the relatives of control subjects, and this association
is grounded in genetics rather than shared familial environment as suggested by adoptive and twin studies
(Siever 1991). The genetics of paranoid personality disorder are less well understood, but it has a high overlap
with schizotypal personality disorder, and its presence
may be greater in families of patients with schizophrenia or delusional disorder (Webb and Levinson 1993).
Schizoid personality disorder has received little or no
genetic study but is more common in the relatives of
patients with schizophrenia (Kalus et al. 1993).

Neuropsychopharmacology
Dopamine System
The dopamine system has been extensively studied in
patients with schizophrenia and particularly associated with the psychotic symptoms of this disorder,
consistent with the antipsychotic effects of the neuroleptics, which act as dopamine antagonists. Accordingly, given the phenomenological and genetic relationships between schizophrenia and schizotypal
personality disorder, the dopaminergic system has
been the primary neurotransmitter system studied in
schizotypal personality disorder.
Neurochemistry. Plasma homovanillic acid (HVA), a
major metabolite of dopamine, has been found to be elevated in clinically selected patients with schizotypal
personality disorder, and this elevation is significantly
correlated with psychotic-like criteria for this disorder,
such that statistical correction for the presence of psychotic-like symptoms abolishes the difference between
groups (Siever et al. 1991). An identical configuration of
results is found with respect to cerebrospinal fluid
(CSF) HVA (Siever et al. 1993). On the other hand,
among relatives of patients with schizophrenia, who
are generally characterized more by the social and cognitive deficit-like symptoms of schizotypal personality
disorder, plasma HVA is lower in subjects with schizotypal personality disorder than control subjects (Amin
et al. 1999). In these studies, plasma HVA was negatively correlated with the negative or deficit-like symp-

Neurobiology

157

toms of schizotypal personality disorder. Interestingly,


however, when the negative symptoms were entered as
a covariant, the positive relationship with psychoticlike symptoms in plasma HVA emerged (Amin et al.
1997). Reduced plasma HVA concentrations have been
associated with impairment in tests of frontally mediated executive function such as the Wisconsin Card
Sort Test (Siever et al. 1991). Thus, these results suggest
that dopaminergic activity may be relatively increased
or decreased depending on the predominance of psychotic-like versus deficit-like symptoms, respectively.
This distinction is consistent with formulations that increased dopaminergic activity, particularly in striatum,
is associated with psychotic-like symptoms and that
decreased dopaminergic activity, particularly in prefrontal regions, is especially associated with deficit-like
symptoms (Siever and Davis 2004).

Longer-term pharmacological interventions have


been evaluated in individuals with schizotypal personality disorder to determine their effects on cognitive function. Preliminary data from studies of guanfacine, an 2-adrenergic agonist, and pergolide, a D1/
D 2 agonist, suggest improvement in cognitive function, particularly working memory, with these catecholaminergic interventions, consistent with the facilitatory effects of the catecholamines on cognitive
function and prefrontal cortex. Cognitive function
may also improve with risperidone (Koenigsberg et al.
2003), possibly due to the effects that 5-HT2 blockade
has on facilitating dopaminergic activity in frontal
lobe. Antipsychotic effects have been documented in a
number of clinical trials of atypical and typical neuroleptics in individuals with schizotypal personality disorder (Hymowitz et al. 1986; Schulz et al. 2003).

Acute Pharmacological Interventions. A mphetamine,


which stimulates the release of the monoamines, particularly dopamine and norepinephrine, has been shown
to improve the cognitive performance of schizotypal
personality disorder subjects on tests of executive function, working memory, and to a lesser extent, sustained
attention and verbal learning (Kirrane et al. 2000; Siegel
et al. 1996). These improvements are more consistent
than those observed in schizophrenic subjects given
amphetamine and are not accompanied by the behavioral worseningthat is, increased psychotic symptomsfound after amphetamine administration in
schizophrenic patients. Indeed, the deficit-like symptoms of schizotypal personality disorder tend to improve following amphetamine administration (Laruelle et al. 2002; Siegel et al. 1996). These results
suggest that agents that enhance catecholamines, including dopamine, may have beneficial effects on cognition, presumably through stimulation of D1 receptors
in prefrontal cortex.
Similarly, the administration of a glucopyruvic stressor, 2-deoxyglucose, which activates stress-sensitive subcortical systems such as the dopamine system and the
hypothalamic-pituitary-adrenal (HPA) axis, results in
greater stress-related (i.e., plasma cortisol and HVA) responses in patients with schizophrenia than in normal
subjects. In contrast, patients with schizotypal personality disorder show normal (plasma HVA) or even reduced
(cortisol) activation compared with control subjects, suggesting that patients with schizotypal personality disorder have better-buffered subcortical stress-responsive
systems than patients with schizophrenia. Consequently,
it is possible that this buffer provides a protective factor
against psychosis in patients with schizotypal personality disorder (Siever and Davis 2004).

DNA Polymorphisms. Catechol-O-methyltransferase


(COMT) plays a critical role in inactivation of dopamine in the frontal lobe, where the dopamine transporter is not the primary mode of inactivation of
dopamine. Recently, a single nucleotide polymorphism in the COMT gene has been discovered. With
this polymorphism, the allele for the COMT gene codes
for the amino acid valine (Val), as opposed to methionine (Met), in the COMT enzyme. The substitution of
Val for Met leads to a COMT enzyme that has far more
activity than a COMT enzyme coded by the MET allele.
Thus, individuals with VAL alleles should have increased activity of COMT compared with those with
the MET allele. Because increased COMT activity is associated with increased destruction of catecholamines,
individuals with VAL alleles should have less central
dopamine activity than those with MET alleles. Consistent with this idea, cognitive impairment, particularly
evident in dopamine-dependent working memory,
has been associated with the presence of the VAL allele
in patients with schizophrenia (Weinberger et al. 2001)
as well as their healthy siblings and control subjects
(Goldberg et al. 2003). Preliminary studies in patients
with schizotypal personality disorder also suggest an
association between cognitive impairment and the
VAL allele, consistent with the role of reduced dopaminergic activity hypothesized to contribute to the
cognitive dysfunction in the schizophrenia spectrum
disorders such as schizotypal personality disorder.

Cognitive Function and Psychophysiology


Although cognitive dysfunction may exist in subtle
forms in a variety of personality disorders, the most
consistent and robust changes are found in people

158

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

with Cluster A personality disorders, more specifically schizotypal personality disorder. Patients with
schizotypal personality disorder show attenuated patterns of cognitive impairment similar to those of patients with schizophrenia but somewhat more specific. For example, overall intelligence may not be
impaired (Mitropoulou et al. 2002; Trestman et al.
1995), whereas specific disturbances in sustained attention, in verbal learning, and particularly in working memory have been reported in patients with
schizotypal personality disorder compared with patients with other non-schizophrenia-related personality disorders, the latter of whom are generally not impaired in these indices, and with normal control
subjects (Mitropoulou et al. 2002). Although patients
with schizophrenia showed deviations from normal
control subjects on the order of two standard deviations, patients with schizotypal personality disorder
have more on the order of one standard deviation below the mean or less (Mitropoulou et al. 2002). The
deficits in working memory and attention may contribute to the impaired rapport and misreading of verbal and facial cues in patients with schizotypal personality disorder, who often clinically complain that they
have a hard time focusing on others, which detracts
from their ability to engage. Indeed, performances on
working memory tasks have been reported to be correlated with interpersonal impairment (Mitropoulou
et al. 2002; Siever et al. 2002).
A variety of psychophysiological endophenotypes
that may reflect genetic substrates to the schizophrenia
spectrum disorders have been found to be abnormal in
patients with schizotypal personality disorder as well
as in patients with chronic schizophrenia. Many of
these psychophysiological abnormalities have also
been found in relatives of patients with schizophrenia,
who may have mild schizophrenia-spectrum symptoms or may even appear to be clinically healthy, raising
the possibility that these abnormalities reflect an underlying genetic susceptibility to the schizophrenia spectrum that is variably expressed. Although a detailed review of psychophysiological abnormalities is beyond
the scope of this chapter, abnormalities in eye movement, visual processing, and inhibition of startle response are among the most consistently replicated.
Thus, individuals with schizotypal personality disorder
showed impaired smooth-pursuit eye movement, antisaccade generation, and velocity discrimination. Furthermore, they show less capacity for inhibition on a
prepulse inhibition paradigm and P50-evoked potential
paradigm. The latter finding is of particular interest because it has been linked to a specific allele of the nico-

tinic receptor in families of patients with schizophrenia.


Backward masking, reflecting early visual processing, has also been reported to be abnormal in patients
with schizotypal personality disorder and schizophrenia (Siever and Davis 2004; see Braff and Friedman
[2002] for an overview of these psychophysiologic abnormalities).

Neuroimaging
(See also Chapter 38, Brain Imaging.)

Structural Imaging
Patients with schizotypal personality disorders show
ventricular enlargement and reduced volumes of several brain regions, as do patients with schizophrenia. In
studies of patients with schizotypal personality disorder, ventricular volume is increased, although studies
of relatives of patients with schizotypal personality disorder are mixed (Shihabuddin et al. 1996; Siever 1995).
Temporal volume reductions in patients with schizotypal personality disorder appear to be comparable
with those observed in schizophrenic patients and occur in both superior temporal gyrus and other temporal
regions. However, some data suggest that frontal volumes are relatively preserved, suggesting that greater
frontal capacity may serve as a buffer against the severe
cognitive and social deterioration we see in schizophrenia. Whereas striatal volumes of patients with schizophrenia are enlarged secondary (in large part) to neuroleptic medications, the striatal volumes (including
putamen [Shihabuddin et al. 2001] and caudate [Levitt
et al. 2002]) of patients with schizotypal personality disorder are reduced in comparison with normal control
subjects and unmedicated patients with schizophrenia.
Reduced striatal volumes are consistent with the possibility of reduced dopaminergic activity, which may be
protective against the emergence of psychosis.

Functional Imaging
Both positron emission tomography (PET) and single
photon emission computed tomography (SPECT) functional imaging studies suggest that patients with
schizotypal personality disorder do not activate regions
such as dorsolateral prefrontal cortex in response to an
executive function or learning task to the same degree
as control subjects, but do so to a greater degree than do
patients with schizophrenia. However, patients with
schizotypal personality disorder are able to activate
other compensatory regions, including the anterior
pole of frontal cortex (Brodmann area 10), which is be-

Neurobiology

lieved to be a high-level executive region (Buchsbaum


et al. 2002). A recent functional magnetic resonance imaging (fMRI) study (Koenigsberg et al. 2001) using a
visuospatial working-memory task also showed increased activation in patients with schizotypal personality disorder in Brodmann area 10, but lesser activation in dorsolateral prefrontal cortex than in normal
control subjects. Thus, patients with schizotypal personality disorder may have compensatory mechanisms
available to them that patients with schizophrenia do
not have in the face of diminished capacity to use dorsolateral prefrontal cortex. The compensatory mechanisms may involve using higher executive regions than
are required for normal individuals.
An IBZM SPECT study measuring dopamine released by displacement of [11C] iodine-methoxybenzamide (IBZM) demonstrated that subjects with
schizotypal personality disorder released significantly
more dopamine in response to amphetamine administration than did normal control subjects but less than
did acute schizophrenic patients (Siever et al. 2002).
These results are consistent with functional imaging
studies suggesting increased activation of ventral striatum, which is normally inhibited by dopamine, in
unmedicated schizotypal patients compared with
control subjects and unmedicated schizophrenic patients as well as the reduced plasma HVA responses to
2-deoxyglucose and striatal volumes noted earlier in
these studies, suggesting dopaminergic activity that is
better buffered than that of schizophrenic patients.

Case Example
Mr. C is a 56-year-old common-law married male,
employed in his extended familys business, whose
current complaint is that people at work are accusing me of saying things that I am not saying. Mr. C
has been seen by the psychiatry service for more
than 20 years, after he was admitted to medicine for
complaints of back pain. He was transferred to psychiatry because he couldnt stand up. He was first
psychiatrically hospitalized when he was in the
Navy for an episode of going crazy after a dispute
with his captain. He had symptoms of depersonalization, irritability, and difficulty getting along with
his peers. Six years after his tour in the Navy, he saw
a therapist but would have vivid dreams that were
disturbing to both the therapist and himself, at
which point Mr. C states his therapy ended. He has
had paranoid ideation, thinking that people at work
are against him, although this suspiciousness and
ideation are responsive to reality testing, as are his
ideas of reference. He has prolonged periods of anhedonia and demoralization but, other than insomnia at times, does not have extensive vegetative
symptoms of depression. He has experienced epi-

159

sodes of depersonalization described as looking


down at himself. He complains of low self-esteem
but denies worthlessness, hopelessness, or helplessness. Mr. C notes that he was always a loner and had
no close friends since the fifth grade. He went to college just before he went to Vietnam. He smokes one
pack of cigarettes per day, does not use recreational
drugs, and drinks up to three drinks per night, although he goes for periods without drinking significantly.
Mr. C underwent a research evaluation in the
Mood and Personality Disorders Program. Research
diagnostic evaluation revealed the presence of a
schizotypal personality disorder with traits of paranoid and narcissistic personality disorder; he was
also found to meet DSM-IV-TR (American Psychiatric Association 2000) criteria for alcohol abuse
(past). Neurobiological evaluation uncovered a
number of abnormalities. First, he displayed modestly impaired eye-movement accuracy (3.38 on a
1=best to 5=worst scale) and mild cognitive impairment. His dopaminergic indices were high, with a
plasma HVA level of 14.5 ng/mL (mean for normal
subjects is 7.4 1.8 ng/mL) and a CSF HVA level of
38.0 ng/mL (mean for normal control subjects = 24.1
6 ng/mL). In addition, Mr. C showed hypofrontality on a PET scan during a verbal memory task. Finally, he showed modest improvement following administration of amphetamine. Since evaluation,
Mr. C has been treated with low-dose neuroleptic
medication that helps him control multiple symptoms, including an olfactory hallucination-like experience of the smell of cordite, a feeling that others
are staring at him, a feeling of being detached or
separated by a bubble from other people, a lack of
any close friends other than a common-law wife,
feelings in the past that his wife might be following
around, and a feeling in the past that he has seen future events.

Summary
These studies suggest that patients with schizotypal
personality disorder have at least a profile of cognitive
impairment and structural brain abnormalities, particularly in temporal cortex, similar to that found in patients with schizophrenia, but a combination of better
prefrontal reserves and more subdued dopaminergic
activity subcortically protect them from the emergence
of psychosis. Their more subtle cognitive impairments
are reflected in their eccentricity and interpersonal disengagement but do not reach the threshold of overt
psychosis. For these reasons, they present more in the
context of their disturbed interpersonal style and coping mechanisms rather than in the context of overt psychosis as in schizophrenia. However, this disorder provides an example of a spectrum that in its more extreme
forms manifests as an Axis I disorder (schizophrenia)

160

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

but in milder forms as an Axis II disorder. There are few


biological data regarding paranoid personality disorder when it is not comorbid with schizotypal personality disorder.

CLUSTER B PERSONALITY DISORDERS


The Cluster B personality disorders include antisocial, borderline, histrionic, and narcissistic personality disorder. Individuals with these disorders present
with varied degrees of impulsivity, aggression, and
emotional dysregulation. As in other clusters, there is
a high degree of overlap among the disorders in Cluster B, particularly between antisocial personality disorder (ASPD) and borderline personality disorder
(BPD). ASPD and BPD are the best studied of the cluster, due to clear and reliable criteria for the former
and the high prevalence of the latter in clinical populations.

Behavioral Genetics
Twin studies suggest that the genetic influence underlying personality disorders is at least as high as that
underlying personality traits that underlie the various
personality disorders. In a relatively small twin study
(Torgersen et al. 2000) that may tend to overestimate
the underlying genetic influence of any of a variety of
personality disorders, the heritability for Cluster B
personality disorders was 0.60. The heritabilities of the
specific Cluster B disorders in this study were 0.79 for
narcissistic personality disorder, 0.69 for BPD, and
0.67 for histrionic personality disorder. The bestfitting models did not include shared familial environment effects, although such effects may influence the
development of BPD. Adoption studies of ASPD confirm a strong genetic, although a less strong environmental, influence for this disorder (Cadoret et al.
1985). Although adoption studies of other Cluster B
personality disorders have not been conducted, the results of family history studies suggest a complex pattern of familial aggregation in which traits related to
impulsive aggressiveness and mood dysregulation,
rather than BPD itself, are transmitted in families (Silverman et al. 1991).

Neuropsychopharmacology
The serotonin (5-HT) system has been extensively
studied in individuals with personality disorder in
general and in particular as an inverse correlate of im-

pulsive aggressive behavior. Other neurotransmitters


and/or modulators have also been studied in this regard, but to a much lesser degree.

Serotonin
There is a clear and consistent role for 5-HT in the regulation of aggression and/or impulsivity, particularly
in individuals with personality disorder. Most data
suggest an inverse relationship between any of a variety of measures of 5-HT levels and levels of aggression
or impulsivity. Although some studies suggest a primary relationship with impulsivity, most studies report a 5-HT relationship more consistent with the construct of impulsive aggression.
Neurochemical Studies. Inverse relationships between
human aggression and measures of central 5-HT
function have been reported since 1979, when Brown
and colleagues reported an inverse relationship between CSF levels of the main central 5-HT metabolite,
5-hydroxyindoleacetic acid (5-HIAA), and life history
of actual aggressive behavior in males with a variety of
DSM-II (American Psychiatric Association 1968) personality disorder diagnoses (Brown et al. 1979). This
finding was extended (Brown et al. 1982) to include a
trivariate relationship between history of aggression,
suicide attempts, and reduced CSF 5-HIAA, whereby
history of aggression and suicide attempts were correlated directly with each other and inversely with CSF
5-HIAA. Later work with violent offenders (Linnoila et
al. 1983) found reduced CSF 5-HIAA in impulsive, but
not nonimpulsive, violent offenders with a variety of
DSM-II personality disorder diagnoses, suggesting that
impulsive aggression was the form most associated
with reduced CSF 5-HIAA concentration. Although
these findings have been replicated, an inverse relationship between CSF 5-HIAA and aggression has not been
reported in samples of individuals with personality disorder without a prominent history of criminal activity
(Coccaro et al. 1997a, 1997b; Gardner et al. 1990; Simeon
et al. 1992). It is likely that CSF 5-HIAA, being a relatively insensitive index of 5-HT activity, is most reduced in the most severely aggressive individuals and
that it is difficult to detect this relationship in less severely aggressive individuals.
Acute Pharmacological Interventions. There are a variety of 5-HT acute pharmacological challenge studies
that have been performed in individuals with personality disorder in the context of the study of aggression.
Typically, hormonal (e.g., prolactin) responses to the
5-HT selective agents are reported to correlate inversely with various measures of aggression and im-

Neurobiology

pulsivity (Coccaro et al. 1989, 1997a, 1997b; Dolan et


al. 2001; Moss et al. 1990; OKeane et al. 1992; Paris et
al. 2004; Siever and Trestman 1993). Pharmacological
challenge studies using putatively receptor-selective
5-HT agents also seem to support the hypothesis of an
inverse relationship between 5-HT and measures of
aggression and suggest a role for at least the 5-HT1A
receptor in particular (Cleare and Bond 2000; Coccaro
et al. 1990, 1995; Hansenne et al. 2002). A more complex picture in regard to central 5-HT1A receptors has
been suggested recently by the observation of reduced
5-HT1A receptormediated responses in females with
BPD with a history of sustained child abuse (Rinne et
al. 2000). Because childhood abuse has been linked to
impulsive aggression in later adolescence and adulthood (Crick and Dodge 1996), it remains to be determined whether the relationships between 5-HT and
aggression are linked to this environmental/developmental variable. Although behavioral responses to
5-HT stimulation in individuals with personality disorder have not received much attention, at least one
study reported a significant reduction in anger in 12
patients with BPD after administration of the mixed
5-HT agonist meta-chlorophenylpiperazine (m-CPP)
but not placebo (Hollander et al. 1994); a reduction in
fear was also observed in the males with BPD.
Platelet Receptor Markers. Despite considerable platelet receptor work in other psychiatric populations, relatively little research in this area has been published
on subjects with personality disorder. Inverse correlations between the number of platelet 3H-imipramine
(5-HT transporter) binding sites and self-mutilation
and impulsivity have been reported in individuals
with personality disorder but not in patients without
a history of self-mutilation (Simeon et al. 1992). Similarly, an inverse correlation between the number of
platelet 3 H-paroxetine (5-HT transporter) binding
sites (Coccaro et al. 1996), and the quantity of platelet
serotonin (Goveas et al. 2004), and life history of aggression has been reported in individuals with personality disorder.
DNA Polymorphism Studies. Work in this area began
with an examination of DNA polymorphisms in the
gene for tryptophan hydroxylase (TPH). TPH is the
rate-limiting step for the synthesis of serotonin, and it
was thought that polymorphisms in TPH would lead
to TPH enzymes of different activities. Although this
TPH polymorphism was not found to have a clear
functional consequence regarding serotonin synthesis,
the presence of the L allele (L referred to the lower
band on the genotyping gel) was found to have some

161

association with clinically relevant variables. For example, impulsive violent offenders (nearly all with a
personality disorder) with at least one copy of the L
TPH allele have been reported to have significantly
lower CSF 5-HIAA compared with impulsive violent
offenders with the UU genotype (U referred to the upper band on the genotyping gel) in at least one study
(Neilson et al. 1994). This finding did not generalize to
nonimpulsive violent offenders (many of whom also
had a personality disorder) or to normal control subjects and was not replicated in a later study by the same
authors (Neilson et al. 1998). The presence of the L allele was associated with an increased risk of suicidal
behavior in all violent offenders in this and in a later
study by these authors (Neilson et al. 1994, 1998). New
et al. (1998) have also reported that the self-reported
tendency toward aggression varies as a function of
TPH genotype whereby subjects with the LL genotype
had higher aggression scores than those with the UU
genotype. Curiously, however, the reverse finding was
reported by Manuck et al. (1999) in a sample of healthy
volunteers from the community: higher aggression
scores were associated with the presence of the U allele. These disparate findings may be due to critical differences in the subject samples. As such, the relationship between the TPH allele and 5-HT function may be
dependent on the TPH alleles relationship with some
other gene depending on the subject sample. Lappalainen et al. (1998) reported an association between
antisocial alcoholism (i.e., alcoholism with ASPD or
intermittent explosive disorder) and the C allele for the
5-HT 1D beta-receptor polymorphism. Because the
5-HT1D beta receptor is a critical receptor involved in
the regulation of 5-HT release on neuronal impulse,
this finding could be highly relevant to the understanding of ASPD comorbid with alcoholism.

Catecholamines
Compared with serotonin, far fewer data have been
published regarding the role of other neurotransmitters and behavioral dimensions of relevance to the
Cluster B personality disorders.
Neurochemical Studies. A positive correlation between
CSF methoxyhydroxyphenylglycol (MHPG, the major
metabolite of norepinephrine) concentrations and life
history of aggression has been reported in males with
personality disorder, although further analysis revealed
that CSF 5-HIAA concentration accounted for most
(80%) of the variance in aggression scores. Similarly,
one study reported a small positive correlation between
plasma norepinephrine and self-reported impulsivity

162

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

in males with personality disorder (Siever and Trestman 1993). In contrast, at least one study (Virkkunen et
al. 1987) reported a significant reduction in CSF MHPG
concentration in males who have committed violent offenses. Finally, a recent study reports an inverse relationship between plasma free MHPG and life history of
aggression in males with personality disorder (Coccaro
et al. 2003). Compared with patients with nonborderline
personality disorders, patients with BPD had lower
plasma free MHPG compared with the nonborderline
control subjects; a finding that disappeared after differences in aggression scores were accounted for. Evidence
for the role of dopamine in aggression in individuals
with personality disorder is limited and contradictory.
Although some studies demonstrate no relationship between CSF HVA concentration and aggression (Brown
et al. 1979; Virkkunen et al. 1987), other studies demonstrate an inverse relationship between these variables
(Linnoila et al. 1983; Virkkunen et al. 1989). Given the
consistent observation of a strong correlation between
CSF 5-HIAA and CSF HVA concentrations, it is possible
that findings with CSF HVA may be related to similar
findings with CSF 5-HIAA concentration. If so, a specific assessment of CSF HVA may not be made unless
the effect of CSF 5-HIAA concentration is accounted for,
a statistical adjustment that has not been made in published studies to date.
Acute Pharmacological Interventions. Early studies of
the acute administration of amphetamine in patients
with BPD demonstrated a greater behavioral sensitivity
to amphetamine challenge among the patients with personality disorder than among control subjects (Schulz et
al. 1985). Replication studies found that global worsening in psychopathology after amphetamine was typical
of patients with both borderline and schizotypal personality disorder whereas global improvement was typical
of borderline subjects without comorbid schizotypal personality disorder (Schulz et al. 1988). This finding suggests important biological differences among patients
with BPD as a function of comorbid schizotypy (perhaps
because of preexisting dopaminergic hyperactivity in
mesolimbic dopamine circuits). In other studies of amphetamine challenge relevant to Cluster B personality
disorder, a direct relationship with affective lability has
been noted in healthy volunteers, suggesting that increases in norepinephrine and/or dopamine may play a
role in the moment-to-moment dysregulation of affect
seen in patients with BPD (Kavoussi et al. 1993). Only
limited data are available regarding the study of norepinephrine receptormediated responses related to the
features of Cluster B personality disorder. One study
reported a positive correlation between the growth hor-

mone response to the 2 norepinephrine agonist clonidine and self-reported irritability (a correlate of aggression) in a small sample of males with personality
disorder and healthy volunteers (Coccaro et al. 1991). A
more recent study of females with BPD, however, reported no difference in growth hormone responses to
clonidine (Paris et al. 2004).
DNA Polymorphism Studies. The presence of the lowfunctioning MAO-A allele in young men combined
with a history of childhood maltreatment has recently
been shown to be associated with an increased risk of
aggressive and criminal offending (e.g., antisocial) behavior (Caspi et al. 2002). This specific MAO-A allele is
associated with reduced catabolism of catecholamines
(and serotonin) and accordingly with higher levels of
these neurotransmitters that may be associated with
aggressive behavior. These data suggest that although
the presence of this allele may be important in increasing the risk of antisocial behavior, the co-occurrence of
childhood maltreatment in vulnerable individuals is
also needed to meaningfully increase the risk of antisocial behavior.

Acetylcholine and Other Neurotransmitters/


Neuromodulators
Studies of acetylcholine function in personality disorder have been limited to two studies. In the first (Steinberg et al. 1997), patients with BPD reported greater
self-rated depression scores in response to the cholinomimetic agent physostigmine than did patients with
nonborderline personality disorders or healthy volunteer control subjects. Peak physostigmine-induced depression scores correlated positively with the number
of affective instability, but not with the number of
impulsive aggression, borderline personality traits.
This finding suggests that the trait of affective lability in
patients with BPD may be mediated in part by a heightened sensitivity to acetylcholine. In the second study
(Paris et al. 2004), however, no differences in hormonal
responses to a different cholinomimetic agent, pyridostigmine, were seen between females with BPD and
control subjects. These divergent findings suggest the
possibility that the cholinergic receptors mediating behavioral and hormonal responses to cholinergic agents
in these subjects may be very different by virtue of
brain location.
Other neurotransmitters or neuromodulators that
may play a role in Cluster Brelated features include
vasopressin, which may have a direct relationship
with aggression (Coccaro et al. 1998); substances related to limbic-hypothalamic-pituitary adrenal axis

Neurobiology

functioning (corticotropin releasing factor, adrenocorticotropic hormone, cortisol), which may have varied
relationships regarding aggressive behavior dependent on social context and stress (Rinne et al. 2002);
testosterone, which is variably correlated with aggression, particularly in violent offenders with ASPD
(Virkkunen et al. 1994); and cholesterol and fatty acids,
which may play a role in both aggression (both: Atmaca et al. 2002; New et al. 1999) and mood regulation
(fatty acids: Zanarini and Frankenberg 2003).

Neuroimaging
(See also Chapter 38, Brain Imaging.)

Structural Imaging
Reduced prefrontal gray matter (e.g., by 11%) has been
associated with autonomic deficits in individuals with
ASPD characterized by aggressive behaviors (Raine et
al. 2000). Conversely, increases in corpus callosum
white matter volume and length have recently been
described in similar subjects (Raine et al. 2003), where
larger callosal volumes were also associated with affective/interpersonal deficit, low autonomic stress reactivity, and spatial ability. Given the complex role
these structures play in mediating cognitive and affective processes, these findings may represent anatomical correlates of the complex behaviors seen in ASPD.
A confounding role for alcoholism in these matters
must always be addressed, however, because it also
has been shown that volume changes may be correlated with duration of alcoholism (Laakso et al. 2002).
Similar structural imaging studies of females with
BPD report reductions in the volume of subcortical
structures such as the amygdala (Rusch et al. 2003;
Schmahl et al. 2003; Tebartz van Elst et al. 2003) and
hippocampus (Schmahl et al. 2003; Tebartz van Elst et
al. 2003). One study also reports reductions in the volumes of both cortical (right orbitofrontal) and other
limbic structures including right anterior cingulated
and amygdala/hippocampal volumes (Tebartz van
Elst et al. 2003). Given the role these structures are
thought to play in emotional information processing,
it is tempting to speculate that these structures represent anatomical correlates of the emotional dysregulation (including impulsive aggression) seen in patients
with BPD.

Functional Imaging (PET and SPECT)


Whereas structural imaging yields only a static picture of the brain, SPECT or PET scanning can yield
functional information related to cerebral blood flow

163

or cerebral glucose metabolism, respectively. For example, SPECT studies have demonstrated reduced
perfusion in prefrontal cortex as well as focal abnormalities in left temporal lobe and increased activity in
anteromedial frontal cortex in limbic system in aggressive individuals with ASPD and alcoholism (Amen et
al. 1996). A more recent study using SPECT reported
significant correlations between reduced cerebral
blood flow in frontal and temporal brain regions and
the disturbed interpersonal attitude factor from the
Psychopathy Checklist (Soderstorm et al. 2002). In homicide offenders (many of whom presumably had
ASPD), a bilateral diminution of glucose metabolism
has been reported in both medial frontal cortex and at
a trend level in orbital frontal cortex (Raine et al. 1994).
In a study of patients with a variety of personality disorders, an inverse relationship was found between life
history of aggressive impulsive behavior and regional
glucose metabolism in orbital frontal cortex and right
temporal lobe (Goyer et al. 1994). Patients meeting criteria for BPD had decreased metabolism in frontal regions corresponding to Brodmann areas 46 and 6 and
increased metabolism in superior and inferior frontal
gyrus (Brodmann areas 9 and 45; Goyer et al. 1994).
More-recent PET studies in females with BPD reported hypometabolism in both frontal and prefrontal
regions as well as in the hippocampus and cuneus
(Juengling et al. 2003), supporting previous structural
studies that demonstrated reductions in the volumes
of these brain areas. Although most of these PET studies were performed in the resting condition, a recent
PET study in females with BPD showed that the replay of abandonment scripts prior to PET scan are associated with greater increases in activity in dorsolateral prefrontal cortex (bilaterally) and in cuneus, but
with reductions in activity in the right anterior cingulate (Schmahl et al. 2003). Given that several of these
structures have been shown to be smaller in these subjects compared with control subjects, the increased activity in these regions after the abandonment task is
quite notable.
PET studies may also be performed after the administration of neurotransmitter-specific agents so
that the activity of brain regions in response to activation of specific receptors by these agents can be assessed. To date, at least four studies of patients with
personality disorder have been performed in this way.
Two utilized the indirect 5-HT agonist fenfluramine,
one utilized the more direct postsynaptic 5-HT agonist
m-CPP, and one examined the trapping of a 11C analogue of tryptophan. In the first fenfluramine study,
patients with prominent histories of impulsive aggres-

164

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

sion and BPD demonstrated blunted responses of glucose metabolism in orbital frontal, ventral medial
frontal, and cingulate cortex compared with normal
subjects (Siever et al. 1999). A similar result was reported in the second fenfluramine study, in which patients with BPD displayed reduced glucose metabolism (relative to placebo) compared with control
subjects in right medial and orbital frontal cortex, left
middle and superior temporal gyri, left parietal lobe,
and left caudate (Soloff et al. 2000). In the PET study
involving m-CPP, patients with prominent histories of
impulsive aggression and personality disorder were
found to have reduced activation of the anterior cingulate and increased activation of the posterior cingulate compared with control subjects (New et al. 2002).
Given the role of the anterior cingulate in emotional
information processing, it is noteworthy that this area
is underactivated by 5-HT stimulation. In the PET
study examining the unilateral trapping of a 11C analogue of tryptophan, evidence for a reduction in 5-HT
synthesis was present in the corticostriatal (e.g., medial frontal, anterior cingulate, superior temporal gyri,
and corpus striatum) brain areas of subjects with BPD
(Leyton et al. 2001). Reduction in 5-HT synthesis in
these regions was reported to correlate with a laboratory measure of behavioral disinhibition.

Functional Imaging (fMRI)


Unlike PET or SPECT, fMRI does not require the injection of a radiolabeled agent. Instead, fMRI assesses
changes in cerebral blood flow using changes in the
blood oxygenation leveldependent signal in the magnetic resonance imaging scanner. This offers a much
greater spatial and temporal resolution compared
with either PET or SPECT and allows a finer assessment of the activation and deactivation of discrete regions of the brain in response to specific stimuli. To
date, at least three studies using fMRI in patients with
personality disorder have been published. In one
study using fMRI, males with (psychopathic) ASPD
activated preselected frontal and temporal regions of
interest less than did control subjects during trials of
negatively charged emotional words (Kiehl et al.
2001), suggesting an important deficit in emotional information processing. In a similar fMRI study in females with BPD, the study group demonstrated
greater activation of the amygdala bilaterally (as well
as activation of selected frontal regions) while viewing
emotionally aversive images (e.g., crying children)
than did control subjects (Herpertz et al. 2001). The
most recent fMRI study in females with BPD reported
a generally similar finding (left amygdala as opposed

to bilateral activation) using emotional faces (Donegan et al. 2003). Given the clear differences in known
emotional information processing between psychopathic antisocial subjects on the one hand and borderline subjects on the other, these data suggest the brain
sites of these differences.

Case Example
Mr. D is a 29-year-old married male computer technician referred for treatment of his impulsive aggressive outbursts in the context of a threatened separation from his wife of 4 years. Mr. D reports impulsive
aggressive outbursts since his mid-teens. These outbursts typically involve screaming, shouting, and
throwing things around; he has only occasionally
physically hit anyone. However, these aggressive
outbursts occur several times a month and usually
several times a week, particularly when Mr. D is
held up in traffic. Most recently, he has been having serious marital difficulty, and his wife is now
threatening to leave him if he does not get help for
his anger problem. He reports that his relationship
with his wife is often stormy, with frequent fighting that sometimes goes on for hours. Sometimes in
the aftermath of these fights Mr. D runs off and gets
exceedingly drunk and drives recklessly around
town while high. At other times, he reports, he beats
his head so hard against a wall that his forehead
bleeds (once he needed stitches). Still, at other times
he frantically pleads with his wife not to leave him;
once he took an overdose of aspirin, in front of his
wife, to get her to stay with him. Mr. D reports a history of alcohol abuse in his late teens and early twenties and a history of gambling to excess up until
1 year prior to evaluation.
Mr. D underwent a research evaluation in the
Mood and Personality Disorders Program. Diagnostic evaluation revealed the presence of BPD with traits
of histrionic, narcissistic, and obsessive-compulsive
personality disorder. He was also found to meet
DSM-IV-TR criteria for two episodes of major depression in the past and for alcohol abuse (past) and
pathological gambling (past). He underwent a variety
of research-related studies including d-fenfluramine
(d-FEN) challenge and was found to have a blunted,
but not absent, prolactin response to d-FEN (2.3 ng/
mL compared with 6.3 3.4 ng/mL for healthy male
control subjects); his CSF 5-HIAA level was not abnormal (23.9 ng/mL compared with 20.0 4.9 ng/mL
for healthy male control subjects). The modest magnitude of his prolactin response to d-FEN suggests a
limited degree of central serotonin system dysfunction. Mr. D entered a treatment trial of fluoxetine and
experienced a reduction in overt aggressive behavior
over a period of several weeks. Over this time his relationship with his wife somewhat improved, and he
is now in dialectical behavioral therapy to work on
other aspects of his interpersonal difficulties with others in his life.

Neurobiology

Summary
The studies discussed in this section suggest that patients with Cluster B personality disorder have dysfunction in a variety of neurobiological areas that may
underlie their clinical presentation. Dysfunction can
occur in multiple monoaminergic systems (e.g., serotonin, norepinephrine, vasopressin for impulsivity
and aggression, possibly acetylcholine for mood reactivity) and in brain structures related to behavioral inhibition and emotional information processing (e.g.,
orbitofrontal cortex, amygdala). Although patients
with borderline personality are often the most extreme in these features and in related biological dysfunction, specific biological dysfunction related to
specific traits (e.g., serotonin dysfunction with impulsive aggression) can be seen in patients with other,
nonborderline personality disorders. As such, it is
doubtful that any assessment of specific neurobiological function will be specific to patients with BPD.

CLUSTER C: ANXIOUS CLUSTER


PERSONALITY DISORDERS
The Cluster C personality disorders include avoidant,
dependent, and compulsive personality disorders. Individuals with these disorders present with varied degrees of anxiety sometimes expressed as rigidity,
particularly in the case of compulsive personality disorder. Of the three disorders, avoidant personality
disorder is most like generalized social phobia in
Axis I, and a great degree of comorbidity occurs between the two diagnoses (Dahl 1996). As in other personality disorder clusters, there is overlap among the
disorders in this cluster and with those in other personality disorder clusters, particularly Cluster B. To
date, there has been much less empirical neurobiological research with patients in Cluster C.

Behavioral Genetics
As with the Cluster B personality disorders, twin studies suggest substantial genetic influence for each of the
Cluster C personality disorders (Torgersen et al. 2000).
Heritability for Cluster C personality disorders as a
group was estimated at 0.62; heritabilities for each disorder in the study were 0.78 for obsessive-compulsive,
0.57 for dependent, and 0.28 for avoidant personality
disorder. The best-fitting models did not include shared
familial environment effects, although a model consisting only of shared familial and unique environmental

165

effects could not be definitively ruled out for dependent


personality disorders. Family studies suggest a familial
association between social anxiety disorder and avoidant personality disorder (Schneier et al. 2002). Avoidant,
dependent, and anxious cluster personality disorders
show significant familiarity (Reich 1989), and both
avoidant and independent personality traits are found
in relatives of patients with panic disorder (Reich 1991).

Neuropsychopharmacology
There has been little biological study of the Cluster C
personality disorders. However, low dopamine metabolites in CSF have been identified in patients with
social anxiety disorder (Johnson et al. 1994), which
overlaps to a great extent with avoidant personality
disorder, whereas nonselective monoamine oxidase
inhibitors (which increase dopamine transmission) or
dopaminergic antidepressants improve social anxiety
(Schneier et al. 2002). Imaging studies are also consistent with this finding, with low dopamine transporter
binding demonstrated in generalized social anxiety
disorder (Tiihonen et al. 1997) and lower D2 receptor
binding in a SPECT study of generalized social anxiety
disorder (Schneier et al. 2000). In addition, three PET
studies support a relationship of reduced D2 binding
associated with detachment, which correlates with social avoidance consistent with that observed both in
patients with Cluster C personality disorders and in
patients with schizoid personality disorder (Schneier
et al. 2000). Genetic studies of these types of behaviors
have been found in association with the dopamine
transporter gene DAT1 (Blum et al. 1997). These studies cumulatively suggest low dopaminergic activity in
social anxiety disorder and likely in avoidant personality disorder as well.
In the serotonergic system, on the other hand, patients with social anxiety have increased cortisol responses to serotonergic agents (Tancer et al. 1999), and
social anxiety disorders respond to selective serotonin
reuptake inhibitors that re-regulate serotonergic activity
(Schneier et al. 2003). Shyness (related to avoidant traits)
has been associated with the serotonin transporter reporter region L allele but not to COMT, MAO-A, or
DRD4 alleles. Growth hormone regulation has also been
associated with social anxiety (Schneier et al. 2002).

Neuropsychologic and Psychophysiologic


Correlates
Increased amygdala activation in fMRI has been
shown in social phobia in one study (Schneier et al.

166

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

1999) as well as in recognition bias for recall of disapproving faces in another (Foa et al. 2000). However,
skin conductance and heart rate change and startle response during viewing of slides with emotionally
charged themes did not distinguish patients with
avoidant personality disorder from control subjects
(Herpertz et al. 2000). Psychophysiologic studies have
not been extensively undertaken in the other Cluster C
personality disorders.

Summary
Genetic and neurobiologic research has been limited in
patients with Cluster C personality disorders but reductions in dopaminergic activity and increases in serotonergic activity are hinted at in the data available.

FUTURE DIRECTIONS
Research in the psychobiology of personality disorder
has advanced much since the 1980s. Although there is
clear evidence of a number of biogenetic correlates of
personality disorder traits, future efforts need to be directed along a variety of lines to increase our understanding of how alterations in brain function lead to
the development and manifestation of these traits.
Such lines of investigation may be aimed at 1) how genetic and environmental influences interact with neurotransmitter function to lead to specific traits; 2) how
neurotransmitter function interacts with the regulation of cognitive and emotional function across distributed neural networks to lead to specific traits; and
3) how understanding brain function at these levels
can enable us to devise more effective ways to treat
personality disorder traits both pharmacologically
and psychotherapeutically.

REFERENCES
Amen DG, Stubblefield M, Carmicheal B: Brain SPECT findings and aggressiveness. Ann Clin Psychiatry 3:129
137, 1996
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 2nd Edition. Washington,
DC, American Psychiatric Association, 1968
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association,
2000

Amin F, Coccaro EF, Mitropoulou V, et al: Plasma HVA in


schizotypal personality disorder, in Plasma Homovanillic Acid Studies in Schizophrenia: Implications for Presynaptic Dopamine Dysfunction. Edited by Friedhoff
AJ, Amin F. Washington, DC, American Psychiatric
Press, 1997, pp 133149
Amin F, Silverman JM, Siever LJ, et al: Genetic antecedents
of dopamine dysfunction in schizophrenia. Biol Psychiatry 45:11431150, 1999
Atmaca M, Kuloglu M, Tezcan E, et al: Serum cholesterol
and leptin levels in patients with borderline personality
disorder. Neuropsychobiology 45:167171, 2002
Blum K, Braverman ER, Wu S: Association of polymorphisms of dopamine D2 receptor (DRD2), and dopamine transporter (DAT1) genes with schizoid/avoidant
behaviors (SAB). Mol Psychiatry 2:239246, 1997
Braff DL, Freedman R: Endophenotypes in studies of the genetics of schizophrenia, in Neuropsychopharmacology:
The Fifth Generation of Progress. Edited by Davis KL,
Charney D, Coyle JT, et al. Philadelphia, PA, Lippincott
Williams and Wilkins, 2002, pp 703716
Brown GL, Goodwin FK, Ballenger JC, et al: Aggression in
humans correlates with cerebrospinal fluid amine metabolites. Psychiatry Res 1:131139, 1979
Brown GL, Ebert MH, Goyer PF, et al: Aggression, suicide,
and serotonin: relationships to CSF amine metabolites.
Am J Psychiatry 139:741746, 1982
Buchsbaum MS, Nenadic I, Hazlett EA, et al: Differential
metabolic rates in prefrontal and temporal Brodmann
areas in schizophrenia and schizotypal personality disorder. Schizophr Res 54:141150, 2002
Cadoret RJ, OGorman TW, Troughton E, et al: Alcoholism
and antisocial personality: interrelationships, genetic
and environmental factors. Arch Gen Psychiatry 42:161
167, 1985
Caspi A, McClay J, Moffitt TE, et al: Role of genotype in the
cycle of violence in maltreated children. Science 297:851
854, 2002
Cleare AJ, Bond AJ: Ipsapirone challenge in aggressive men
shows an inverse correlation between 5-HT1A receptor
function and aggression. Psychopharmacology 148:344
349, 2000
Coccaro EF, Siever LJ, Klar HM, et al: Serotonergic studies in
affective and personality disorder: correlates with suicidal and impulsive aggressive behavior. Arch Gen Psychiatry 46:587599, 1989
Coccaro EF, Gabriel S, Siever LJ: Buspirone challenge: preliminary evidence for a role for 5-HT-1A receptors in
impulsive aggressive behavior in humans. Psychopharmacol Bull 26:393405, 1990
Coccaro EF, Lawrence T, Trestman R, et al: Growth hormone
responses to intravenous clonidine challenge correlates
with behavioral irritability in psychiatric patients and
in healthy volunteers. Psychiatry Res 39:129139, 1991

Neurobiology

Coccaro EF, Kavoussi RJ, Hauger RL: Physiologic responses


to d-fenfluramine and ipsapirone challenge correlate
with indices of aggression in males with personality
disorder. Int Clin Psychopharmacol 10:177180, 1995
Coccaro EF, Kavoussi RJ, Sheline YI, et al: Impulsive aggression in personality disorder: correlates with 3H-Paroxetine binding in the platelet. Arch Gen Psychiatry
53:531536, 1996
Coccaro EF, Kavoussi RJ, Cooper TB, Hauger RA. Central
serotonin and aggression: inverse relationship with prolactin response to d-fenfluramine, but not with CSF
5-HIAA concentration in human subjects. Am J Psychiatry 154:14301435, 1997a
Coccaro EF, Kavoussi RJ, Trestman RL, et al: Serotonin function in personality and mood disorder: intercorrelations
among central indices and aggressiveness. Psychiatry
Res 73:114, 1997b
Coccaro EF, Kavoussi RK, Hauger RL, et al: Cerebrospinal
fluid vasopressin: correlates with aggression and serotonin function in personality disordered subjects. Arch
Gen Psychiatry 55:708714, 1998
Coccaro EF, Lee R, McCloskey M: Norepinephrine function
in personality disorder: plasma free MHPG correlates
inversely with life history of aggression. CNS Spectr
8:731736, 2003
Crick NR, Dodge KA: Social information-processing mechanisms in reactive and proactive aggression. Child Dev
67:9931002, 1996
Dahl AA: The relationship between social phobia and
avoidant personality disorder: workshop report 3. Int
Clin Psychopharmacol 11(suppl 3):109112, 1996
Dolan M, Anderson IM, Deakin JF: Relationship between
5-HT function and impulsivity and aggression in male
offenders with personality disorders. Br J Psychiatry
178:352359, 2001
Donegan NH, Sanislow CA, Blumberg HP, et al: Amygdala
hyperreactivity in borderline personality disorder: implications for emotional dysregulation. Biol Psychiatry
54:12841293, 2003
Foa EB, Gilboa-Schechtman E, Amir N: Memory bias in generalized social phobia: remembering negative emotional expressions. J Anxiety Disord 14:501519, 2000
Gardner DL, Lucas PB, Cowdry RW: CSF metabolites in borderline personality disorder compared with normal
controls. Biol Psychiatry 28:247-54, 1990
Goldberg TE, Egan MF, Gscheidle T, et al: Executive subprocesses in working memory: relationship to catechol-Omethyltransferase Val158Met genotype and schizophrenia. Arch Gen Psychiatry 60:889896, 2003
Goveas JS, Csernansky JG, Coccaro EF: Platelet serotonin
content correlates inversely with life history of aggression in personality-disordered subjects. Psychiatry Res
126:2332, 2004
Goyer PF, Andreason PJ, Semple WE, et al: Positron-emission tomography and personality disorders. Neuropsychopharmacology 10:2128, 1994

167

Hansenne M, Pitchot W, Pinto E, et al: 5-HT1A dysfunction


in borderline personality disorder. Psychol Med 32:935
941, 2002
Herpertz SC, Schwenger UB, Kunert HJ, et al: Emotional responses in patients with borderline as compared with
avoidant personality disorder. J Personal Disord 14:
339351, 2000
Herpertz SC, Dietrich TM, Wenning B, et al: Evidence of abnormal amygdala functioning in borderline personality
disorder: a functional MRI study. Biol Psychiatry 50:
292298, 2001
Hollander E, Stein D, DeCaria CM, et al: Serotonergic sensitivity in borderline personality disorder: preliminary
findings. Am J Psychiatry 151:277280, 1994
Hymowitz P, Frances A, Jacobsberg LB, et al: Neuroleptic
treatment of schizotypal personality disorders. Compr
Psychiatry 27:267271, 1986
Johnson MR, Lydiard RB, Zealberg JJ: Plasma and CFS levels
in panic patients with comorbid social phobia. Biol Psychiatry 36:426427, 1994
Juengling FD, Schmahl C, Hesslinger B, et al: Positron emission tomography in female patients with borderline
personality disorder. J Psychiatr Res 37:109115, 2003
Kalus O, Bernstein DP, Siever LJ: Schizoid personality disorder: a review of its current status. J Personal Disord 7:
4352, 1993
Kavoussi RJ, Coccaro EF: The amphetamine challenge test
correlates with affective lability in healthy volunteers.
Psychiatry Res 48:219228, 1993
Kiehl K, Smith AM, Hare RD, et al: Limbic abnormalities in
affective processing by criminal psychopaths as revealed by functional magnetic resonance imaging. Biol
Psychiatry 50:677684, 2001
Kirrane RM, Mitropoulou V, Nunn M, et al: Effects of amphetamine on visuospatial working memory performance in schizophrenia spectrum personality disorder.
Neuropsychopharmacology 22:1418, 2000
Koenigsberg HW, Buchsbaum MS, Harvey P, et al: Regional
brain activation in schizotypal personality disorder patients during visuospatial working memory task as
measured by fMRI. Abstract of the Fifty-sixth Annual
Meeting of the Society of Biological Psychiatry, Vol 49,
abstract 327, 2001
Koenigsberg HW, Reynolds D, Goodman M, et al: Risperidone in the treatment of schizotypal personality disorder. J Clin Psychiatry 64:628634, 2003
Laakso MP, Gunning-Dixon F, Vaurio O, et al: Prefrontal volumes in habitually violent subjects with antisocial personality disorder and type 2 alcoholism. Psychiatry Res
114:95102, 2002
Lappalainen J, Long JC, Eggert M, et al: Linkage of antisocial
alcoholism to the serotonin 5-HT1B receptor gene in
two populations. Arch Gen Psychiatry 55:989994, 1998
Laruelle M, Kegeles L, Zea-Pance Y, et al: Amphetamineinduced dopamine release in patients with schizotypal
personality disorders studies by SPECT and [123]
IBZM. Neuroimage 16:S61, 2002

168

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Levitt JJ, McCarley RW, Dickey CC, et al: MRI study of caudate nucleus volume and its cognitive correlates in neuroleptic-nave patients with schizotypal personality disorder. Am J Psychiatry 159:11901197, 2002
Leyton M, Okazawa H, Diksic M, et al: Brain regional alpha[11C]methyl-L-tryptophan trapping in impulsive subjects with borderline personality disorder. Am J Psychiatry 158:775782, 2001
Linnoila M, Virkkunen M, Scheinin M, et al: Low cerebrospinal fluid 5-hydroxyindolacetic acid concentration differentiates impulsive from nonimpulsive violent behavior. Life Sci 33:26092614, 1983
Manuck SB, Flory JD, Ferrell RE, et al: Aggression and angerrelated traits associated with a polymorphism of the
tryptophan hydroxylase gene. Biol Psychiatry 45:603
614, 1999
Mitropoulou V, Harvey PD, Maldari LA, et al: Neuropsychological performance in schizotypal personality disorder: evidence regarding diagnostic specificity. Biol Psychiatry 52:11751182, 2002
Moss HB, Yao JK, Panzak GL: Serotonergic responsivity and
behavioral dimensions in antisocial personality disorder with substance abuse. Biol Psychiatry 28:325338,
1990
New AS, Gelernter J, Yovell Y, et al: Tryptophan hydroxylase
genotype is associated with impulsive aggression measures. Am J Med Genet 81:1317, 1998
New AS, Sevin EM, Mitropoulou V, et al: Serum cholesterol
and impulsivity in personality disorders. Psychiatry
Res 85:145150, 1999
New AS, Hazlett EA, Buchsbaum MS, et al: Blunted prefrontal cortical 18fluorodeoxyglucose positron emission tomography response to meta-chlorophenylpiperazine in
impulsive aggression. Arch Gen Psychiatry 59:621629,
2002
Nielsen DA, Goldman D, Virkkunen M, et al: Suicidality and
5-hydroxyindoleacetic acid concentration associated
with a tryptophan hydroxylase polymorphism. Arch
Gen Psychiatry 51:3438, 1994
Nielsen DA, Virkkunen M, Lappalainen J, et al: A tryptophan hydroxylase gene marker for suicidality and alcoholism. Arch Gen Psychiatry 55:593602, 1998
OKeane V, Moloney E, ONeill H, et al: Blunted prolactin responses to d-fenfluramine in sociopathy: Evidence for
subsensitivity of central serotonergic function. Br J Psychiatry 160:643646, 1992
Paris J, Zweig-Frank H, Kin NM, et al: Neurobiological correlates of diagnosis and underlying traits in patients
with borderline personality disorder compared with
normal controls. Psychiatry Res 121:239252, 2004
Raine A, Buchsbaum M, Stanley J, et al: Selective reductions
in prefrontal glucose metabolism in murderers. Biol
Psychiatry 36:365373, 1994
Raine A, Lencz T, Bihrle S, et al: Reduced prefrontal gray matter volume and reduced autonomic activity in antisocial
personality disorder. Arch Gen Psychiatry 57:119127,
2000

Raine A, Lencz T, Taylor K, et al: Corpus callosum abnormalities in psychopathic antisocial individuals. Arch Gen
Psychiatry 60:11341142, 2003
Reich JH: Familiality of DSM-III dramatic and anxious personality clusters. J Nerv Ment Dis 177:96100, 1989
Reich JH: Avoidant and dependent personality traits in relatives of patients with panic disorder, patients with dependent personality disorder, and normal controls. Psychiatry Res 39:8998, 1991
Rinne T, Westenberg HG, den Boer JA, et al: Serotonergic
blunting to meta-chlorophenylpiperazine (m-CPP)
highly correlates with sustained childhood abuse in impulsive and autoaggressive female borderline patients.
Biol Psychiatry 47:548556, 2000
Rinne T, de Kloet ER, Wouters L, et al: Hyperresponsiveness
of hypothalamic-pituitary-adrenal axis to combined
dexamethasone/corticotropin-releasing hormone challenge in female borderline personality disorder subjects
with a history of sustained childhood abuse. Biol Psychiatry 52:11021112, 2002
Rusch N, van Elst LT, Ludaescher P, et al: A voxel-based
morphometric MRI study in female patients with borderline personality disorder. Neuroimage 20:385392,
2003
Schmahl CG, Vermetten E, Elzinga BM, et al: Magnetic resonance imaging of hippocampal and amygdala volume
in women with childhood abuse and borderline personality disorder. Psychiatry Res 122:193198, 2003
Schneier F, Weiss U, Kessler C: Subcortical correlates of differential classical conditioning of aversive emotional reactions in social phobia. Biol Psychiatry 45:863871,
1999
Schneier FR, Liebowitz MR, Abi-Dargham A: Low dopamine D2 receptor binding potential in social phobia.
Am J Psychiatry 157:457459, 2000
Schneier FR, Blanco C, Smita XA, et al: The social anxiety
spectrum. Psychiatr Clin North Am 25:757774, 2002
Schneier FR, Blanco C, Campeas R, et al: Citalopram treatment of social anxiety disorder with comorbid major
depression. Depress Anxiety 17:191196, 2003
Schulz SC, Schulz PM, Dommisse C, et al: Amphetamine response in borderline patients. Psychiatry Res 15:97108,
1985
Schulz SC, Cornelius J, Schulz PM, et al: The amphetamine
challenge test in patients with borderline personality
disorder. Am J Psychiatry 145:809814, 1988
Schulz SC, Thomson R, Brecher M: The efficacy of quetiapine vs. haloperidol and placebo: a meta-analytic study
of efficacy. Schizophr Res 62:112, 2003
Shihabuddin L, Silverman JM, Buchsbaum MS, et al: Ventricular enlargement associated with linkage marker for
schizophrenia-related disorders in one pedigree. Mol
Psychiatry 1:215222, 1996
Shihabuddin L, Buchsbaum MS, Hazlett EA, et al: Striatal
size and relative glucose metabolic rate in schizotypal
personality disorder and schizophrenia. Arch Gen Psychiatry 58:877884, 2001

Neurobiology

Siegel BV, Trestman RL, OFlaithbheartaigh SO, et al:


D-amphetamine challenge effects on Wisconsin Card
Sort Test performance in schizotypal personality disorder. Schizophr Res 20:2932, 1996
Siever LJ: The biology of the boundaries of schizophrenia, in
Advances in Neuropsychiatry and Psychopharmacology, Vol 1: Schizophrenia Research. Edited by Tamminga CA, Schulz SC. New York, Raven, 1991, pp 181
191
Siever LJ: Brain structure/function and the dopamine system in schizotypal personality disorder, in Schizotypal
Personality. Edited by Raine A, Lencz T, Mednick F.
New York, Cambridge University Press, 1995, pp 272
286
Siever LJ, Davis KL: The pathophysiology of the schizophrenic disorders: perspective from the spectrum. Am
J Psychiatry 161:398413, 2004
Siever LJ, Trestman RL: The serotonin system and aggressive
personality disorder. Int Clin Psychopharmacol 8 (suppl
2):3339, 1993
Siever LJ, Amin F, Coccaro EF, et al: Plasma homovanillic
acid in schizotypal personality disorder patients and
controls. Am J Psychiatry 148:12461248, 1991
Siever LJ, Amin F, Coccaro EF, et al: CSF homovanillic acid
in schizotypal personality disorder. Am J Psychiatry
150:149151, 1993
Siever LJ, Buchsbaum M, New A, et al: D,L-fenfluramine response in impulsive personality disorder assessed with
[18]-fluorodeoxyglucose positron emission tomography. Neuropsychopharmacology 20:413423, 1999
Siever LJ, Koenigsberg HW, Harvey P, et al: Cognitive and
brain function in schizotypal personality disorder.
Schizophr Res 54:157167, 2002
Silverman JM, Pinkham L, Horvath TB, et al: Affective and
impulsive personality disorder traits in the relatives of
patients with borderline personality disorder. Am J Psychiatry 148:13781385, 1991
Simeon D, Stanley B, Frances A, et al: Self-mutilation in personality disorders: psychological and biological correlates. Am J Psychiatry 149:221226, 1992
Soderstrom H, Hultin L, Tullberg M, et al: Reduced frontotemporal perfusion in psychopathic personality. Psychiatry Res 114:8194, 2002
Soloff PH, Meltzer CC, Greer PJ, et al: A fenfluramine-activated FDG-PET study of borderline personality disorder. Biol Psychiatry 47:540547, 2000

169

Steinberg BJ, Trestman R, Mitropoulou V, et al: Depressive


response to physostigmine challenge in borderline personality disorder patients. Neuropsychopharmacology
17:264273, 1997
Tancer ME, Mailman RB, Stein MB, et al: Subcortical correlates of differential classical conditioning of aversive
emotional reactions in social phobia. Biol Psychiatry
45:863871, 1999
Tebartz van Elst L, Hesslinger B, Thiel T, et al: Frontolimbic
brain abnormalities in patients with borderline personality disorder: a volumetric magnetic resonance imaging study. Biol Psychiatry 54:163171, 2003
Tiihonen J, Kuikka J, Bergstrom K: DA reuptake site densities in patients with social phobia. Am J Psychiatry
154:239242, 1997
Torgersen S, Lygren S, Oien PA, et al: A twin study of personality disorders. Compr Psychiatry 41:416425, 2000
Trestman RL, Keefe RSE, Harvey PD, et al: Cognitive function and biological correlates of cognitive performance
in schizotypal personality disorder. Psychiatry Res
59:127136, 1995
Virkkunen M, Nuutila A, Goodwin FK, et al: Cerebrospinal
fluid monoamine metabolite levels in male arsonists.
Arch Gen Psychiatry 44:241247, 1987
Virkkunen M, DeJong J, Bartko J, et al: Relationship of psychobiological variables to recidivism in violent offenders and impulsive fire setters. Arch Gen Psychiatry 46:
600603, 1989
Virkkunen M, Rawlings R, Tokola R: CSF biochemistries,
glucose metabolism, and diurnal activity rhythms in alcoholic, violent offenders, fire setters, and healthy volunteers. Arch Gen Psychiatry 51:2027, 1994
Webb CT, Levinson DF: Schizotypal and paranoid personality disorder in the relatives of patients with schizophrenia and affective disorders: a review. Schizophr Res
11:8192, 1993
Weinberger DR, Egan MF, Bertolino A, et al: Prefrontal neurons and the genetics of schizophrenia. Biol Psychiatry
50:825844, 2001
Zanarini MC, Frankenburg FR: Omega-3 Fatty acid treatment of women with borderline personality disorder: a
double-blind, placebo-controlled pilot study. Am J Psychiatry 160:167169, 2003

This page intentionally left blank

11
Developmental Issues
Patricia Cohen, Ph.D.
Thomas Crawford, Ph.D.

Aside from well-documented developmental links between early conduct disorder and antisocial personality disorder (ASPD), there are large gaps in our
knowledge about childhood antecedents of other
DSM-IV-TR personality disorders (American Psychiatric Association 2000; Widiger and Sankis 2000). In
this chapter we discuss how developmental processes
and selected risk factors lead to the emergence and
persistence of personality disorders in young people.
We highlight changes in how children and adolescents construct mental representations of themselves
and other people and then consider how distortions in
this developmental process manifest in personality
disorder. We outline what we have learned about
early trajectories of personality disorders and discuss
how clinicians and researchers can evaluate the normative and clinical significance of symptoms in children and adolescents. Finally, we address problems in
assessing these disorders in young people based on
the limited number of measurement instruments currently available.

ETIOLOGICAL AND DEVELOPMENTAL


FACTORS
In an early paper on the borderline-child-to-be, Pine
(1986) identified three key factors in his developmental model of borderline personality disorder (BPD) in
young people. First, he hypothesized how early abuse
or trauma overwhelms the child, especially when the
trauma is ongoing or experienced from a variety of
sources. Second, childhood trauma may interfere with
the development of how trust, libidinous attachments,
anxiety, aggression, and self-esteem are experienced
and expressed. Third, young people may fasten onto
immature defenses almost as though these defenses
were survival techniques for desperate situations.
More recent empirical research points to other factors
that contribute to the emergence and persistence of
personality disorders over time, including genetic effects and co-occurring Axis I disorders (Crawford et
al. 2001b; Livesley et al. 1998).
171

172

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

When individual risk factors occur in isolation,


they often may be offset by normative maturational
factors in social or cognitive domains. Conduct disorder in childhood, for instance, does not usually lead
to ASPD in adulthood. When risk factors occur in
combination, however, they may overwhelm the
young persons ability to cope, thus leading immature defenses to become inflexible and maladaptive
over time. On the other hand, even a child who has
experienced a significant trauma may be protected
from lasting damage to personality functioning if he
or she is securely attached to parents who can buffer
the impact of the trauma. If traumatized children are
anxiously attached to parents instead, thus reducing
the protective effects, they may be at greater risk for
lasting personality dysfunction (Alexander 1992). Because genetic, interpersonal, and early trauma risk
factors are all addressed elsewhere in this textbook
(see Chapter 9, Genetics; Chapter 12, Attachment
Theory and Mentalization-Oriented Model of Borderline Personality Disorder; and Chapter 13, Role of
Childhood Experiences in the Development of Maladaptive and Adaptive Personality Traits), we focus
here on disturbances in how children perceive themselves and the people around them and how developmental changes in cognitive ability may play a role in
the formation of personality disorders. In this context
we draw on the theoretical literature on attachment in
infancy and childhood (e.g., Cassidy and Shaver 1999;
Fonagy et al. 2003) and identity development in adolescence (Erikson 1968) and seek to bridge the two
using Harters (1998) work on development of selfrepresentation.

Early Working Models of Self and Other


Attachment theory (Bowlby 1969, 1973) focuses on developmental experiences reflecting secure and insecure relationships between infants and caregivers and
emphasizes how young people come to perceive
themselves and others. A secure attachment typically
occurs when the caregiver has been available and sensitive to the needs of the infant or toddler, especially in
times of distress. Young children can better manage
negative emotions, such as anger or fear, within a secure relationship because these feelings have been associated with soothing and effective responses by the
caregiver (Sroufe 1996). By providing this external
form of affect regulation, caregivers prevent infants
from being overwhelmed and help them gradually
develop the ability to regulate their own affect. When
caregivers are inconsistent or rejecting, infants and

toddlers instead may underregulate their own affect


or restrict it excessively.
As hypothesized in attachment theory, very basic
mental representations of self and others are thought to
emerge during infancy through affective experiences
characterizing the childcaregiver relationship. These
preverbal experiences are labeled internal working
models and broadly reflect whether infants expect
caregivers to be available or helpful when needed.
Young children may also internalize a basic sense of
whether they are worthy of love and whether other people can be trusted to provide love and emotional support. Individuals who have predominately negative
self-representations usually have anxious attachment
styles, and those who have predominantly negative representations of others tend to have avoidant attachment
styles. Avoidant attachment is thought to stem from
cool, rejecting, and distant treatment by attachment figures, and anxious attachment is traced to inconsistent
and unpredictable treatment by early attachment figures (Ainsworth et al. 1978; Rothbard and Shaver 1994).
As a guide to behavior, internal working models
influence whether young children seek to regulate affective distress by approaching or by avoiding attachment figures, or even by alternating between these opposing strategies for managing negative emotions.
With growth in cognitive capacity, these basic mental
representations of self and other (schemas) are subject to elaboration, refinement, and increasing differentiation from affective experiences that occur in close
relationships. Nevertheless, these schemas appear to
have a remarkably enduring impact on interpersonal
strategies used to regulate emotional distress (e.g.,
Waters et al. 2000). Developmental changes often reflect heterotypic continuity in how attachment styles
are expressed in different relationships across developmental stages. That is, a negative self-schema may
generate anxious preoccupation with changing attachment figuresa primary caregiver in early childhood,
a peer group in adolescence, a romantic partner in
adulthoodbut nevertheless reflect the same basic
difficulty regulating affect across the different relationships.
Although distorted working models accompanying
insecure attachment are not pathological by themselves, they nevertheless may contribute to the formation of Axis II psychopathology, especially when combined with other risks or biological vulnerability.
Markedly negative representations of others, for instance, may explain higher levels of distrust and suspiciousness in avoidant preadolescents when compared
with more securely attached age peers. From this devel-

Developmental Issues

opmental starting point, a variety of pathways leading


to pathological and nonpathological outcomes may depend on heritable and environmental risk factors that
also influence how personality unfolds. A negative
working model of others, reinforced by a hostile and secretive family environment and combined with a biological vulnerability to Cluster A disturbances, may
foster a developmental trajectory leading toward paranoid or schizotypal disturbances in adolescence and
adulthood. In the absence of biological vulnerability,
negative schemas regarding others and corresponding
behaviors may gradually be modified as young people
learn that their family is not typical of the broader social
environment. Normal maturation processes thus may
reduce the likelihood that early Cluster A disturbances
persist over time.
Early maltreatment by caregivers may produce a
serious disturbance called disorganized attachment (Solomon and George 1999). Abused infants and children
often experience sharp conflicts when approaching
caregivers for comfort and support when they also expect maltreatment from themthus provoking unstable fluctuations between conflicting attachment strategies and behavior. Furthermore, children may fear
the loss of the caregivers they depend on, thus limiting
their ability to experience or express any age-appropriate anger or aggression toward that person. This
phenomenon probably pertains most to the development of the marked instability in interpersonal relationships associated with borderline psychopathology
(Fonagy et al. 2000).

Developmental Changes in How the


Self and Others Are Perceived
Harter (1998) described how cognitive development in
infancy, childhood, and adolescence leads to changes
in how young people experience their sense of self.
Harter characterized the self as a cognitive structure
around which behavior is organized, thus anchoring it
squarely within the larger framework of personality.
As a product of the interaction of biological and social
forces, the self undergoes progressive change throughout development. Despite these changes, the self provides a sense of continuity and a source for scripts to
organize behavior, thereby creating a foundation for
later identity.
Harter emphasized how self-representation often
reflects self-evaluation, a process of comparing oneself
with other people or with an ideal self, which evolves
over time as new cognitive abilities emerge during development. Self-evaluation may be filled with inflated

173

self-worth at one end of the spectrum or laden with selfcontempt at the other, and both ends of the spectrum
may play an early role in the formation of personality
disorders. Even at these extremes, however, cognitive
development may nevertheless allow gradual movement toward a more accurate self-representation with a
balanced integration of positive and negative attributes. Although not addressed in self-perception literature, per se, changes in cognitive development probably influence how others are perceived in an analogous
manner.
One aspect of the cognitive development of the self
can be seen in how children describe themselves at different ages. At early ages children typically describe
themselves by their physical characteristics, typical
behaviors, or material possessions. Self-perception
thus lacks much coherence or integration and self-representation is organized instead around all-or-none
thinking (all good or all bad). Because negative and
positive characteristics are polar opposites, the child
cannot recognize that a single person can have both.
Given their inability to distinguish real and ideal
selves, young children typically have unrealistically
positive self-perceptions that often shade into childhood grandiosity. Older children can admit to negative characteristics in one domain while retaining a
positive self-representation in another. Vacillation between positive and negative self-image in early adolescence is gradually replaced by a more integrated
sense of self and a greater awareness of the importance
of the context to behavior. Early grandiosity thus subsides as young people gain the ability to integrate conflicting self-perceptions into a coherent whole in adolescence and early adulthood.
Inaccurate but age-appropriate self-perceptions in
young children thus may resemble later symptoms of
narcissistic personality disorder. If unrealistically positive self-representations become inflexible and persist
over time, they may limit the young persons ability to
abandon immature self-representations as their cognitive resources and perceptual skills increase during the
course of normal development. When serving defensive functions against childhood adversity, early grandiosity may persist and harden into personality disorder symptoms. Furthermore, grandiosity may be
pathological when asserted aggressively as a way to
prevent all good self-representations from shifting
and suddenly becoming all bad. This defensive style
may lead young people with narcissistic disturbances
to have dismissing or derogatory perceptions of others.
In BPD, young people may lack sufficient internal defenses to prevent self-representations from alternating

174

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

frequently between the extremes of all good and all


bad. Their perception of others similarly alternates between extremes of idealization and devaluation, thus
constituting the clinical phenomenon called splitting.
Among severely abused children, negative selfperceptions may predominate over positive selfimages and lead those children to feel profoundly
unworthy and unlovable (Fischer and Ayoub 1994). In
abusive family environments, caregivers typically reinforce negative evaluations of the child that are then
incorporated into the childs self-representations. As a
result, there may be little foundation for any cognitive
structure of self that would allow the child to develop
and integrate both positive and negative self-evaluations. Furthermore, negative self-evaluations may become automatic (Siegler 1991) in ways that make them
even more resistant to change.
Formal operational thinking, including logical and
abstract reasoning abilities, normally emerges in adolescence (Keating 1990). Dramatic increases in differentiation between self and other also occur during adolescence (Bowlby 1973) and thus increase young
peoples ability to view themselves as distinct from
caregivers. Because differentiation facilitates greater
autonomy, dependence on parental attachment figures normally declines during adolescence as young
people instead identify more with peer groups. At
present we know little about how peer relationships
influence personality maturation during adolescence,
but this important social factor may reinforce self-perceptions that are more internally based and less centered around the parentchild relationship. Although
elevated dependency might not necessarily be pathological earlier in adolescence, it may become a symptom of dependent personality disorder if it persists
past late adolescence and on into early adulthood.
The advent of formal operational thinking provides adolescents with a greater capacity to evaluate
and compare their relationships with different attachment figures, not just with one another but also
against hypothetical ideals. The adolescents ability to
consider attachment relationships in the abstract may
bring with it recognition that parents are deficient in
some ways (Kobak and Cole 1994), perhaps provoking
a dismissive rejection of the parents or angry preoccupation with their shortcomings. Gains in adolescent
insight into parentchild relationships will ideally
lead to greater openness, objectivity, and flexibility as
young people reevaluate attachment relationships
and attain a more realistic and integrated perception
of parents. This developmental process may reduce
dismissing behaviors or dramatic outbursts of anger

that might appear earlier in adolescence to be symptoms of narcissism or histrionic personality disorder. If
parents respond to these adolescent behaviors in maladaptive ways based on their own interpersonal disturbances, they may not facilitate the resolution of
parentchild conflicts and may instead reinforce narcissistic and histrionic tendencies past developmental
stages when they normally decline. Linehan (1993)
emphasized how invalidating responses from parents
can contribute substantially to the formation of BPD,
and Bezirganian et al. (1993) have documented that
kind of relationship empirically.

Self-Understanding, Self-Direction, and


Identity
Erikson (1968) argued that once young people gain
greater awareness of themselves and more accurate
perceptions of others, they often experience a normative crisis of identity during adolescence and early
adulthood. This identity crisis is one of eight age-specific normative crises that occur in human development from infancy through old age. According to
Erikson, a crisis is a turning point when development
must move in one direction or another. In adolescence,
young people either move toward consolidating a secure and stable sense of self or they experience diffuse
identities that provide limited direction or sense of
continuity over time. When the developmental crisis
of identity is successfully resolved, it normally leads
to increased integration of personality. When an identity crisis goes unresolved, it may result in potentially
pathological delays in maturation instead.
Identity consolidation during adolescence primarily entails establishing a clear sense of self and finding a place in the community. Identity is broadly defined in Eriksons theory to encompass self-esteem,
satisfaction with personal and occupational goals, and
confidence in coping skills. Group membership and
sexual identity represent other important domains in
a young persons identity. Identity diffusion, on the
other hand, is typically expressed in the inability to select clear occupational goals or the adoption of roles
deviating from conventional social norms. Erikson
notes that identity diffusion may at times include delinquent behavior or psychotic-like symptoms but
cautions that these disturbances are often transient
during adolescence. Any significant disturbances in
social and emotional development may act to distort
or interfere with normative identity consolidation,
perhaps thereby contributing to the persistence of
early personality disorder symptoms that otherwise

Developmental Issues

might resolve through normal maturational processes.


As defined by Erikson, identity diffusion shares
many characteristics with Axis II symptoms (Cloninger
et al. 1993; Kernberg 1975; Taylor and Goritsas 1994). Indeed, identity disturbances are explicitly included in diagnostic criteria for BPD. Identity disturbances are
strongly implied in the suggestibility to other peoples
influence in histrionic personality disorder, idealized
but unrealistic self-perceptions in narcissistic personality disorder, marked worry about other peoples criticism in avoidant personality disorder, and difficulties in
making everyday decisions in dependent personality
disorder. Moreover, identity diffusion and personality
disorder symptoms share similar developmental trajectories: both decline with age during adolescence and
early adulthood (Johnson et al. 2000a; Meeus et al. 1999).
Distinctions between normal identity diffusion and
more enduring personality disorder symptoms in adolescence may be difficult to make in clinical evaluations.
Erikson noted that it is normal for young people to try
on different identities during adolescence and later
abandon them when they fail to fit comfortably with
their sense of identity. For instance, early attempts to express sexuality may manifest in provocative dress during adolescence without necessarily being a symptom
of histrionic personality disorder. After trying out that
overt expression of sexuality, young people may subsequently opt for less provocative attire that corresponds
more with their internal sense of self. On the other
hand, if provocative dressing co-occurs with poorly
regulated affect and maladaptive preoccupation with
interpersonal relationships, it may signal a more lasting
disturbance of personality.
Despite the broad overlap between personality disorder symptoms and Eriksons construct of identity diffusion, relatively little research has investigated the association between the two. It thus remains unclear
whether identity disturbances contribute to the emergence of personality disorders in adolescence and later
persistence into adulthood or if personality disorder
symptoms delay the consolidation of identity. Consistent with Eriksons epigenetic theory of development,
Cluster B symptoms do appear to interfere with the formation of lasting and committed romantic relationships
that represent the key developmental task of early
adulthood (Crawford et al. 2004). Identity consolidation
may occur at a critical stage in the development of personality disorders because it coincides with a period
when parental influence declines and youths increasingly assert their independence. As young people gradually separate from the family, the identity they choose
plays that much greater a role in defining their person-

175

ality. If adolescents are unable to clearly differentiate


themselves from their parents or to resolve any ongoing
disturbances in their relationship, they are likely to carry
internalized versions of those difficulties with them in
how they perceive themselves and others, perhaps even
recapitulating those disturbances in new relationships
in adulthood. In other words, if identity remains poorly
differentiated as young people separate from their family, any corresponding interpersonal disturbances may
become self-perpetuating during adulthood.

PERSONALITY MATURATION AND AXIS II


PSYCHOPATHOLOGY
Personality traits reflect a complex adaptive system to
internal and environmental conditions, including
changes in affective and cognitive structures during development (Caspi 1998). Specific affects, behaviors, and
cognitions that are age appropriate or normative at one
stage of personality development may reflect immaturity or psychopathology at subsequent ages. As young
people gain emotional and cognitive skills, they usually
abandon immature ways of experiencing and interacting with the world around them. On the other hand,
when young people continue to experience affects, behaviors, or cognitions that their peers have outgrown,
they may encounter interpersonal difficulties that in
some cases accumulate over time. Interpersonal difficulties may be traced to various deficits in the development
of affect regulation during infancy, the formation of conscience during early childhood, the establishment of
age-appropriate impulse control in childhood and adolescence, or the consolidation of identity in late adolescence and early adulthood. Although deficits or delays
in emotional development do not necessarily signify
Axis II pathology, they may indicate that an individual
is on a deviant pathway with increased risk for further
maladaptive behavior. Persistence on a deviant pathway is related to increasing difficulty in returning to a
more normal developmental trajectory. Maturational
change remains possible, but given the organizational
function of personality, developmental change will be
constrained by the individuals previous history. (For a
further discussion of developmental considerations, see
Geiger and Crick 2001 and Kernberg et al. 2000.)

Emotional and Behavioral Problems and


Personality Disorder Symptoms
Children and adolescents appear to outgrow many
problem behaviors that are reflected in current symp-

176

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

tom criteria for personality disorder diagnoses. For instance, parent reports on the Child Behavior Checklist
(CBCL; Achenbach 1991) show a significant linear
decline in withdrawn behavior from age 4 to age 18
(Bongers et al. 2003). Withdrawn behavior pertains not
just to Cluster A personality disorders (paranoid, schizoid, and schizotypal personality disorders) but also to
avoidant personality disorder. Parent reports on individual CBCL items associated with Cluster B personality disorders (borderline, histrionic, and narcissistic personality disorders and conduct disorder that precedes
ASPD) similarly indicate declines in bragging, showing
off, demanding attention, getting into fights, lying,
cheating, having a hot temper, crying a lot, feeling excessive dependence, having problems with peers, and experiencing jealousy (Achenbach 1991). Despite broad
normative reductions in emotional and behavioral
problems, parents report average increases in some
childhood and adolescent symptoms such as being suspicious, secretive, and obsessively preoccupied with
certain thoughts. Furthermore, age changes in symptoms in normative samples do not necessarily follow the
same trajectories as children and adolescents brought in
for clinical evaluation or treatment (Achenbach 1991).
Parents may take normative age changes into account
when assessing the well-being of their children and thus
become concerned only when the expected normative
decline does not appear, at least for some problems.
It may be useful to put these changes in symptom
levels in the context of changes in the prevalence of
Axis I symptoms and disorders over childhood and
adolescence. Disruptive behavior shows a standard
inverted U-shaped prevalence distribution in a wide
range of studies (see Moffitt 1993 for a review), with
large increases from childhood to adolescence and a
sharp drop in young adulthood. Based on parent reports, there are different curvilinear trajectories for
mean levels of anxious and depressive symptoms for
boys and girls over the full age range from 418 years,
with higher rates of problems for boys in childhood
followed by an adolescent decline (Bongers et al. 2003).
For girls these problems increase until the transitional
stage of puberty and then level off. Using teachers as
informants in a large national epidemiological study,
McDermott (1996) found means on every symptom
cluster changing with age between ages 5 and 17 years,
often quite differently for males and females.

Normal and Abnormal Personality Traits


Many researchers view personality disorder symptoms as extreme variants of personality traits that are

continuously distributed in the population (Costa and


Widiger 2002; Livesley et al. 1998). We currently have
only partial information on developmental changes in
trait levels and no information in childhood because of
problems in conceptualizing and measuring personality before adolescence (Shiner and Caspi 2003).
Perhaps the most complete information available
on age changes is based on the five-factor model (FFM)
and its component facets that attempt to measure variation in the full normative range of personality. These
broad dimensions and more narrowly defined facets
have also been proposed as a way of understanding
and potentially measuring personality disorder (Costa
and Widiger 2002; see also Chapter 3, Categorical and
Dimensional Models of Personality Disorders). The
dimension labeled neuroticism is especially pertinent
because it is hypothesized to reflect many of the criteria
for personality disorder. Age changes in self-reported
scores on the Revised NEO Personality Inventory
(NEO-PI-R; Costa and McCrae 1992) were recently
evaluated in a longitudinal sample of gifted children at
ages 12 and 16 years and in a much larger cross-sectional sample of Flemish adolescents between 14 and 18
years (McCrae et al. 2002). Most age changes were very
small and did not follow previously established adult
trajectories that show a gradual decline in neuroticism
from the college years to age 30 (e.g., McCrae et al.
1999). When evaluated in younger adolescent samples,
neuroticism showed an elevation in girls up to about
age 14 and stability thereafter, and there were no significant age changes in boys. Within each broader factor of
the FFM, individual facetage correlations were sometimes different in direction in this study, with some
neuroticism facets increasing with age, some stable
over the age span, and some declining.
Although knowledge of normative age changes in
personality disorder symptoms is limited, the available data clearly show that nearly every Axis II disorder has a gradual linear decrease in the mean number
of symptoms between ages 10 and 25 (Johnson et al.
2000a). Based on combined maternal and child reports,
Figure 111 depicts normative declines in mean symptom levels for approximately 800 youths in the Children in the Community random sample (Cohen and
Cohen 1996), which has been studied longitudinally
since 1975. Mean scores shown indicate the presence of
symptoms without saying anything about how scores
are dispersed. Individual scores in the clinical range,
for instance, fall well above mean values displayed in
Figure 111. When mean levels are higher overall,
more youths are likely to meet fixed diagnostic criteria
that do not take changing age norms into account,

Developmental Issues

thereby possibly increasing the rate of false-positive diagnoses during adolescence. These data suggest that
age-specific norms may be desirable for an assessment
instrument in this age range. However, they also make
clear that the problem of changing normative symptom levels is not limited to childhood and adolescence.
Despite changes in symptom level norms, adolescent
psychiatric disorders warrant clinical attention even in
developmental periods when they are most prevalent.
In keeping with this realization, it is important to determine the association of childhood and adolescent
personality disorder with impairment and negative
prognosis, as is discussed later.
On the whole, normative data do not show any clear
congruence between average age changes in normal
personality dimensions measured on the NEO-PI-R
and normative changes observed using personality disorder measures based on combined youth and parent
reports or parent-reported measures of clinically relevant emotional and behavior symptoms. Some of the
discrepancy may reflect how the NEO-PI-R assesses
personality traits in the normal range and may be limited by ceiling effects at the extreme range of functioning assessed by personality disorder measures or the
CBCL. Once again, the direction of age-related changes
in symptoms in normative community samples may
not always correspond to age-related change in symptoms in children from clinical populations.

CHANGE AND STABILITY IN CHILDHOOD


Given our knowledge that the prevalence of particular
behaviors changes with age, what can be said about the
correlational or rank-order stability of personality or
its temperamental precursors over childhood and adolescence? In particular, to what extent may we expect
that the same individuals who manifest the most extreme personality problems at one age will be among
those who do so at another age?

Temperament
Often regarded as an early precursor to personality,
temperament reflects basic biological differences in
childhood characteristics such as activity level, fearful
withdrawal, ability to be soothed, responsiveness to
stimuli, and affective intensity. Temperament is typically measured by observational ratings or maternal
report. Cloninger and his colleagues have developed a
measure assessing temperament dimensions in preschool children that are posited to be related to later

177

personality disorder (Constantino et al. 2002).


In general, temperament shows significant but low
stability in early childhood (Rothbart and Bates 1998)
for reasons that may be intrinsic to the developmental
process. For example, there may be effects specific to
maturational levels due to genetic or other constitutional influences. The impact of contextual factors may
vary at different maturational levels. Constitutional
environmental interactions also may be a more important source of variation in early childhood, when strong
behavioral habits have not yet become firmly established. There may be more error in observation-based
measures of temperament in infants and preschool children than in older children due to their greater reactivity to fatigue, hunger, and other temporary influences.
Increased measurement error thus contributes to lower
stability estimates in younger children.
An additional problem has been an absence of consensus on how to define the major temperament dimensions, an issue that has only recently begun to be
resolved. Some dimensions of temperament and the
measures devised to assess them have been theoretically derived (Rothbart et al. 2001; Tellegen 1985), including predicted relationships with specific personality disorders (Cloninger 1987; Cloninger et al. 1993).
However, theoretically derived dimension names
sometimes do not clearly correspond to the content of
items. Despite an array of unique construct-derived
names, certain dimensions of temperament can be
viewed in the frame of the FFM (Shiner and Caspi
2003). This frame has the advantage of uniting the personality and temperament literatures but does only
partial justice to the original conceptions of the important individual differences in early childhood such as
executive control and emotional reactivity. Research
linking temperament measured in early childhood to
later personality disorder is only beginning to appear
(Constantino et al. 2002).

Personality Dimensions
There is clear evidence that, on average, a personality
dimension assessed by a self-report instrument will
show lower correlation over equivalent time for
younger persons than for older persons (Roberts and
DelVecchio 2000). Because instability in personality
dimensions continues throughout life, stability coefficients reach a maximum at about age 50 and remain
far below the reliabilities of the measures even then.
Stabilities for symptom measures of personality disorder are very likely to show a similar pattern. Although
we have no preadolescent data, correlations measur-

178

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

25

Mean number of symptom criteria

20

15

10

Total PD traits
Cluster A traits
Cluster B traits

Cluster C traits

0
912

1316

1720

2124

2528

Age, years

Figure 111.
Source.

Declining personality disorder (PD) symptoms by age.

Adapted from Johnson et al. 2000a.

ing stability are moderate in adolescence (Crawford et


al. 2001a), and it is not clear that they increase from adolescence into young adulthood (Johnson et al. 2000a).
General stability may not be the issue if disorders are
outcomes of geneenvironment interactions, so that it
is expected that there may be lower stability on extreme scores than over the full range of relevant dimensions (DiLalla et al. 2000).
Axis II diagnoses, on the other hand, tend to be
much less stable in clinical and community samples of

adolescents (Bernstein et al. 1993;Mattanah et al. 1995).


Axis II diagnoses in adults similarly tend to be unstable
even in clinical samples (Shea et al. 2002; Zanarini et al.
2003). This instability is likely to be a consequence not
only of the generally poorer measurement quality of dichotomized continuous measures (MacCallum et al.
2002) but also of somewhat arbitrary decisions about
diagnostic cut-points. Despite the relative instability of
categorically defined diagnoses, there is increasing evidence of long-term impairment and poor prognosis as-

Developmental Issues

sociated with adolescent personality disorder or high


personality disorder symptoms independent of Axis I
disorders or symptoms (Crawford et al. 2001b; Johnson
et al. 1999, 2000b; Kasen et al. 1999; Lofgren et al. 1991;
Rey et al. 1997). We are thus accumulating evidence that
the criteria for adult disorders may be useful indicators
of Axis II pathology even in young samples and not
necessarily less predictive at times when they are more
normative. However, there are no current studies large
enough to investigate these issues for individual diagnostic criteria, and thus at present this question remains
unanswered at the criterion level.

Case Examples
Given what we know about developmental trajectories of personality disorder symptoms in adolescents,
it appears unwise to make categorical diagnoses during initial clinical assessment except perhaps in extreme cases. Nevertheless, it is meaningful to consider
youths at risk for subsequent personality disorder
based on how they present for treatment and based on
collateral risk factors in close family members, as illustrated by the following vignette.
A 15-year-old boy from an intact family was brought
for individual psychotherapy to address uncontrolled anger, a pervasive hatred toward age peers,
and oppositional and self-defeating behavior (e.g.,
threatening to drop out of high school). His selfimage reflected grandiosity or self-contempt depending on different social contexts, and he tended
to be avoidant and dismissive of others. During
treatment it became evident that one parent tried to
cope with excessive anxiety by becoming overinvolved in the patients long-term plans in ways that
provoked stubborn and oppositional behavior. The
other parent had recurrent depressive episodes, an
explosive and unpredictable temper, identity diffusion, and a dismissive interpersonal style.
On initial evaluation, this 15-year-old youth appeared at increased risk for lasting BPD based on
persistent symptoms of uncontrolled anger, marked
antipathy for others, and poorly integrated representations of self and others. The long-term risk may be
compounded by the presence of parental anxiety
and mood disturbances, especially as they are woven into parentchild relationships. Personality disturbances thus appeared to be reinforced by defensive reactions to one parents overinvolvement and
also by identification with the other parents angry
and dismissive interpersonal style. Despite poor social adjustment with peers, this youth nevertheless
reported having a positive relationship with his parents. Given this protective factor and an absence of
key risk factors such as childhood trauma, his personality disorder symptoms may well subside over

179

time as he gains maturity. Treatment in this case focused on assisting the youth to regulate angry affect
more adaptively and to articulate how peers upset
him so much. Treatment also helped the parents to
become more aware of how their own anxiety and
mood disturbances contributed to maladaptive parentchild interactions in ways that inhibited the
youths gradual movement toward independent
adult functioning.

Just as maturational factors appear to confound


the assessment of personality disorders in adolescents, they may also obscure the presence of personality disorder in adults who have outgrown some earlier
manifestations of the disorder. Knowledge of normative trajectories of personality disorder thus can inform the assessment and treatment of personality disturbances in adult patients whose symptoms fall short
of current diagnostic criteria:
A 42-year-old woman presented for treatment with
complaints of loneliness and an enduring inability
to establish a stable romantic relationship. She recalled adolescence and early adulthood as periods
characterized by labile mood, frequent abandonment fears, volatile and unpredictable romantic
relationships, reactive anger toward parents and
peers, and a series of suicidal gestures. During
childhood she witnessed violent conflicts between
mother and father before they divorced and experienced significant emotional neglect afterward.
When presenting for treatment, this patient denied
any mood lability or suicidal ideation and functioned well at work. However, abandonment fears,
reactive anger, and mood lability recurred whenever she became involved in new romantic relationships, thereby prompting a general avoidance of
close relationships and reinforcing unwanted feelings of loneliness.

Even allowing for inaccuracies that distort retrospective clinical reports, this patient undoubtedly met
full criteria for BPD during adolescence and early
adulthood. Although her symptoms have since fallen
below diagnostic threshold, either due to maturational factors or the effects of prior individual psychotherapies, she nevertheless continues to experience
significant borderline psychopathology in ways that
become painfully apparent whenever new romantic
attachments evoke long-standing abandonment fears
she otherwise seeks to avoid. Treatment in this case focused on clarifying and then reducing abandonment
fears and addressing how her reactive anger undermined the stability of her romantic relationships.
Treatment also addressed the disorganizing effects of
childhood trauma and neglect.

180

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

WHAT IS THE BEST MEASUREMENT


STRATEGY FOR ASSESSING EARLY
PERSONALITY DISORDERS?
Three overall issues pose problems for the assessment
of personality disorders in children and adolescents.
As indicated earlier, DSM-IV-TR does not take into account normative developmental changes in the prevalence of certain problems and the consequent possibility that certain diagnostic criteria should not be seen
as abnormal behavior at some ages. Second, there is a
shortage of appropriate and validated diagnostic instruments for children and adolescents and unresolved questions about the best measurement strategy. Third, aspects of personality are less stable in
childhood and adolescence, thus raising concern that
early diagnosis of personality disorders might lead to
premature labeling.

Choice of Informant
In clinical assessment of adults the patient is usually
the primary informant, although the corroboration and
independent perspectives of knowledgeable sources
such as family members are welcome and often useful.
Research measures designed for use with adults in
clinical or other samples generally are confined to selfreport. In the child mental health field, there is still little consensus about which informants should be considered primary and which should be seen as auxiliary
at different points in childhood development. It is often assumed that the primary caretaker should be the
principal informant for preschool children. As such, instruments measuring temperament, personality, and
psychopathology in children younger than age 9 or 10
years are most often based on information supplied by
parentsfor example, the Childrens Behavior Questionnaire (Rothbart et al. 2001), the Personality Inventory for Children (PIC-2; Lachar 1999a), and the CBCL.
For young children in preschool or day care, teachers
or childcare leaders may provide the best data based
on greater familiarity with normative patterns of behavior for the age.
Similarly, parents and teachers are both thought to
provide relevant data for older children in elementary
school, although in general, agreement on the childs
problems is poor across these informants (Achenbach
et al. 1987). Sometime in childhood, at least by age 9 or
10, most children are able to provide data on their own
characteristics and problems. For instance, the youthreported version of the Diagnostic Interview Schedule

for ChildrenIV (Shaffer et al. 2000) assesses psychopathology in children as young as 9 years old. Similarly, the Personality Inventory for Youth (PIY; Lachar
and Gruber 1995) uses self-reports to assess personality in children starting at the same age. Most selfreport measures of normal personality, however, are
used only with adolescents and adults (Shiner and
Caspi 2003).
Agreement between parent and offspring on youth
behaviors and problems is not very good (e.g., Achenbach et al. 1987), and age changes in prevalence based
on youth self-report look different from those based on
parent report or teacher report. Agreement between self
and informant reports on personality disorder measures for adults is similarly poor (Klonsky et al. 2002).
Despite these difficulties, the research field has reached
general consensus that all informants add usefully to
the assessment of Axis I disorders in children and adolescents. As a consequence, the preferred strategy is to
obtain data from at least two informants and consider
any symptomatic report to be valid providing there is
evidence of associated impairment. If only a single informant can be used, evidence indicates that adolescents are better informants about emotions and often
acknowledge disruptive or antisocial behaviors that
may be unknown to the parent. Parents or teachers are
thought to be better informants on issues where normative comparisons are relevant.
Where does this leave us when deciding which informants to assess for personality disturbances in children and adolescents? At present we do not know
whether diagnostic criteria for Axis II are intrinsically
more difficult to assess in youth than most criteria for
Axis I. All things considered, it appears prudent to
gather data from multiple sources whenever possible
when assessing children and adolescents.

RESEARCH AND CLINICAL MEASURES OF


CHILD OR ADOLESCENT PERSONALITY
DISORDER
As interest in early Axis II disturbances in young people has grown, a variety of instruments designed to
measure normal and abnormal personality in adults
have been evaluated for use in adolescent samples.
Certain personality disorder instruments for adults
have been specifically adapted to be age appropriate
for child or adolescent respondents. Among self- and
parent-reported instruments, Axis II scales are typically combined with various measures of Axis I dis-

Developmental Issues

turbances and thus facilitate the assessment of cooccurrence between these psychiatric constructs. On
the whole, individual scales for these instruments
have internal consistency reliability comparable with
or favorable to adult instruments recently evaluated
by Clark and Harrison (2001).
The largest full assessment of Axis II disorders in a
general community sample of children was undertaken in the Children in the Community Study (Cohen
and Cohen 1996) long before any Axis II scales were
developed for children or adolescents. Prior to assessing this large sample of youths (ranging in age between 9 and 19 years) and their mothers in 1983, Drs.
Cohen and Kasen selected relevant items from scales
included in the research protocol that corresponded to
DSM-III (American Psychiatric Association 1980) diagnostic criteria and then added items or adapted
items from then-current adult instruments to assess
diagnostic criteria not already covered elsewhere
(Bernstein et al. 1993). After data collection was complete, a team of clinical researchers reviewed these
youth- and parent-reported items and wrote algorithms for Axis II symptom scales and categorical diagnoses. Drs. Schwab-Stone and Cohen updated this
assessment by adding items relevant to DSM-III-R
(American Psychiatric Association 1987) for a followup assessment 2.5 years later at mean age 16 (Bernstein et al. 1993). A new set of DSM-IV (American Psychiatric Association 1994) algorithms was developed
to be consistent across these two waves of data collection and an additional assessment of the sample 6
years later at a mean age of 22 (Johnson et al. 1999,
2000a). Validity information on these Axis II scales and
diagnoses comes from a series of analyses of the longitudinal data showing prediction of both long-term impairment and dysfunction independently of Axis I
disorders (Bernstein et al. 1993, 1996; Johnson et al.
1999, 2000b; Kasen et al. 2001). Despite the utility of
this research instrument, it has only recently been
adapted for use by other researchers and is not designed to be a clinical instrument.

Structured Clinical Interviews


Of the structured interviews designed to assess DSMdefined personality disorders in adults, the Personality Disorder Evaluation (Loranger 1988) has been
most thoroughly evaluated and appears to be a valid
measure of Axis II disturbance in adolescents. Nevertheless, more work is needed to identify age-related
differences in adolescent and adult manifestations of
personality disorders. In a longitudinal comparison of

181

adolescent and adult inpatient samples, personality


disorders assessed with the Personality Disorder Evaluation were less stable over a 2-year interval in adolescents than in adults (Mattanah et al. 1995). However,
threshold effects often add unreliability to stability estimates of categorically defined personality disorders,
thus making comparisons across age groups more difficult to interpret. When assessed as dimensional constructs, stability estimates for personality disorders
usually appear higher than when assessed as categorical constructs.

Self-Report Instruments
Although self-report instruments are easier and more
cost-efficient to administer than structured interviews,
questions are raised about whether respondents have
sufficient self-awareness or willingness to acknowledge Axis II symptoms that might stigmatize them.
Given problems in setting reliable thresholds, the available instruments for children and adolescents tend to
assess Axis II symptoms using continuous scales instead of making formal diagnoses.
The Millon Adolescent Clinical Inventory (MACI)
is a well-known instrument modeled on the Millon
Clinical Multiaxial Inventory designed for adults (see
Davis et al. 1999). Intended for adolescents as young
as age 13, the MACI uses 160 self-report items to measure personality disorder constructs congruent with
DSM-defined personality disorders but also reflecting
Millons (1990) theory of personality. The MACI thus
measures 12 personality styles labeled Introversive,
Inhibited, Doleful, Submissive, Dramatizing, Egotistic, Unruly, Forceful, Conforming, Oppositional, SelfDemeaning, and Borderline Tendency. Computer-generated scores on the MACI make adjustments for age
and gender differences in Axis II disturbances based
on norms from separate samples of normal and disturbed adolescents. Standardized scores are further
adjusted to take estimated base rates of psychopathology into account even though the prevalence of Axis II
disturbances in adolescents has yet to be established.
These built-in adjustments are not readily transparent
and effectively preclude their use for investigating the
population prevalence of adolescent Axis II disturbances or for assessing developmental change.

Parallel Parent- and Youth-Reported


Instruments
The PIC-2 and PIY were both originally modeled on
the Minnesota Multiphasic Personality Inventory and

182

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

thus do not correspond directly to DSM-IV Axis II disorders. The parent-reported PIC-2 uses 275 forcedchoice items to measure constructs labeled Cognitive
Impairment, Impulsivity and Distractibility, Delinquency, Family Dysfunction, Reality Distortion, Somatic Concern, Psychological Discomfort, Social
Withdrawal, and Social Skills Deficits. The youthreported PIY uses 270 forced-choice items to measure
the same constructs. As suggested by the labels, the
PIY and PIC-2 measure constructs that probably tap a
mixture of Axis I and Axis II disturbances. Despite
substantial similarity on the PIY and PIC-2, youth and
parent informants show moderate agreement (median
correlation = 0.43, range 0.280.53) (Lachar 1999b).
These concordance rates appear better than the 0.25
correlation between youth and parent reports for comparable age groups on the CBCL (Achenbach et al.
1987).

Additional Instruments for DSM-IV/


DSM-IV-TR Personality Disorders
The Adolescent Psychopathology Scale (Reynolds
1998) is a self-report measure designed for adolescents
between 12 and 19 years old. It measures five of the ten
DSM-IV personality disorders (borderline, avoidant,
obsessive-compulsive, paranoid, and schizotypal personality disorders) along with conduct disorder as the
childhood precursor of ASPD. The Coolidge Personality and Neuropsychological Inventory for Children
(CPNI; Coolidge 1998) assesses DSM-IV Axis II disorders in children and adolescents from 517 years old
by parent report. The CPNI assesses symptoms of all
DSM-IV personality disorders as well as conduct disorder symptoms. The Shedler-Westen Assessment
Procedure200 for Adolescents (SWAP-200-A; Westen
et al. 2003) is a newly developed Q-sort instrument designed for use by skilled clinical observers to assess
Axis II pathology in adolescent patients they see in
treatment. Q-sort is a method by which items are arrayed by the clinician from most descriptive of the adolescent to least descriptive of the adolescent. This
measure was adapted from the Shedler-Westen Assessment Procedure200, a Q-sort designed for adults
that has shown evidence of validity, reliability, and
utility in taxonomic research with adult samples (e.g.,
Shedler and Westen 1998). At present none of the
available instruments has demonstrated clear superiority in clinical and research applications, and there
are limited validity data available for newly developed scales measuring DSM-IV and DSM-IV-TR personality disorders.

CONCLUSIONS
During the past 15 years there has been accumulating
evidence that clinically meaningful personality disorders occur in adolescents. Adolescent personality disorders are associated with emotional distress and psychosocial impairment in community samples (Golombek et
al. 1987; Lewinsohn et al. 1997; Marton et al. 1987; Stein
et al. 1987) and clinical samples in inpatient and outpatient settings (Brent et al. 1994; Grilo et al. 1996; Pinto et
al. 1996; Westen et al. 2003). When adolescent and adult
personality disorders are compared, many similarities
have been observed. Just as adult personality disorders
are associated with co-occurring Axis I disturbances
(Bienvenu and Stein 2003; Farmer and Nelson-Gray
1990), adolescent personality disorders also have been
shown to co-occur with Axis I disturbances (Becker et
al. 2000; Crawford et al. 2001a). However, more work is
needed on age-related differences in adolescent and
adult manifestations of personality disorders.
Although elevated personality disorder scores in
adolescence represent a risk for subsequent psychiatric and psychosocial disturbances, early personality
disorder symptoms are also likely to decline over
time. Caution is thus warranted in evaluating their
clinical significance during adolescence. On the other
hand, there is almost universal agreement that prevention of mental disorders is best accomplished at a very
young age and in collaboration with parents, particularly when other risk factors are also present. Therefore, children with elevated symptoms or their precursors are prime targets for secondary preventionthat
is, for prevention of further developmental delays or
elevation to frank disorder. Regardless of how childhood and adolescent personality disorders are definedeither as an early version of adult disorder or
an early indicator of elevated risk of adult disorder
it appears appropriate to consider interventions to alleviate Axis II disturbances when they manifest in
young people.

REFERENCES
Achenbach TM: Manual for the Child Behavior Checklist,
Ages 418 and 1991 Profile. Burlington, VT, University
of Vermont, Department of Psychiatry, 1991
Achenbach TM, McConaughy SH, Howell CT: Child/adolescent behavioral and emotional problems: implications of
cross-informant correlations for situational specificity.
Psychol Bull 101:213232, 1987

Developmental Issues

Ainsworth MDS, Blehar MC, Waters E, et al: Patterns of attachment: a psychological study of the Strange Situation. Hillsdale, NJ, Erlbaum, 1978
Alexander PC: Application of attachment theory to the
study of sexual abuse. J Consult Clin Psychol 60:185
195, 1992
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition. Washington,
DC, American Psychiatric Association, 1980
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 3rd Edition, Revised. Washington, DC, American Psychiatric Association, 1987
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition. Washington,
DC, American Psychiatric Association, 1994
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision.
Washington, DC, American Psychiatric Association,
2000
Becker DF, Grilo CM, Edell WS, et al: Comorbidity of borderline personality disorder with other personality disorders in hospitalized adolescents and adults. Am J Psychiatry 157:20112016, 2000
Bernstein DP, Cohen P, Velez CN, et al: Prevalence and stability of the DSM-III-R personality disorders in a community-based survey of adolescents. Am J Psychiatry
150:12371243, 1993
Bernstein DP, Cohen P, Skodol A, et al: Childhood antecedents of adolescent personality disorders. Am J Psychiatry 153:907913, 1996
Bezirganian S, Cohen P, Brook JS: The impact of mother
child interaction on the development of borderline personality disorder. Am J Psychiatry 150:18361842, 1993
Bienvenu OJ, Stein MB: Personality and anxiety disorders: a
review. J Personal Disord 17:139151, 2003
Bongers IL, Koot HM, van der Ende J, et al: The normative
development of child and adolescent problem behavior.
J Abnorm Psychol 112:179192, 2003
Bowlby J: Attachment and Loss, Vol 1: Attachment. New
York, Basic Books, 1969
Bowlby J: Attachment and Loss, Vol 2: Separation. New
York, Basic Books, 1973
Brent DA, Johnson B, Perper J, et al: Personality disorder,
personality traits, impulsive violence and completed
suicide in adolescents. J Am Acad Child Adolesc Psychiatry 33:10801086, 1994
Caspi A: Personality development across the life course, in
Handbook of Child Psychology, Vol 3: Social, Emotional, and Personality Development, 5th Edition. Edited by Damon W, Eisenberg N. New York, Wiley, 1998,
pp 311388
Cassidy J, Shaver PR: Handbook of Attachment: Theory, Research, and Clinical Applications. New York, Guilford,
1999
Clark LA, Harrison JA: Assessment instruments, in Handbook of Personality Disorders. Edited by Livesley WJ.
New York, Guilford, 2001, pp 277306

183

Cloninger CR: A systematic method for clinical description


and classification of personality variants: a proposal.
Arch Gen Psychiatry 44:573588, 1987
Cloninger CR, Svrakic DM, Przybeck TR: A psychobiological model of temperament and character. Arch Gen Psychiatry 50:975990, 1993
Cohen P, Cohen J: Life Values and Adolescent Mental Health.
Mahwah, NJ, Lawrence Erlbaum, 1996
Constantino JN, Cloninger CR, Clarke AR, et al: Application
of the seven-factor model of personality to early childhood. Psychiatr Res 109:229244, 2002
Coolidge FL: Coolidge Personality and Neuropsychological
Inventory for Children Manual: CPNI. Colorado Springs,
CO, FL Coolidge, 1998
Costa PT Jr, McCrae RR: Revised NEO Personality Inventory
(NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI)
Professional Manual. Odessa, FL, Psychological Assessment Resources, 1992
Costa PT Jr, Widiger TA (eds): Personality Disorders and the
Five-Factor Model of Personality, 2nd Edition. Washington, DC, American Psychological Association, 2002
Crawford TN, Cohen P, Brook JS: Dramatic erratic personality
disorder symptoms, I: continuity from early adolescence
into adulthood. J Personal Disord 15:319335, 2001a
Crawford TN, Cohen P, Brook JS: Dramatic erratic personality disorder symptoms, II: developmental pathways
from early adolescence to adulthood. J Personal Disord
15:336350, 2001b
Crawford TN, Cohen P, Johnson JG, et al: The course and
psychosocial correlates of personality disorder symptoms in adolescence: Eriksons developmental theory
revisited. J Youth Adolesc 33:373387, 2004
Davis RD, Woodward M, Goncalves A, et al: Treatment planning and outcome in adults: the Millon Clinical Multiaxial Inventory-III, in The Use of Psychological Testing
for Treatment Planning and Outcomes Assessment, 2nd
Edition. Edited by Maruish ME. Mahwah, NJ, Erlbaum,
1999, pp 10511081
DiLalla DL, Gottesman II, Carey G: Madness beyond the
threshold? Associations between personality and psychopathology, in Temperament and Personality Development Across the Life Span. Edited by Molfese
VJ, Molfese DL. Mahwah, NJ, Erlbaum, 2000, pp 177
210
Erikson EH: Identity: Youth and Crisis. New York, WW
Norton, 1968
Farmer R, Nelson-Gray RO: Personality disorders and depression: hypothetical relations, empirical findings, and
methodological considerations. Clin Psychol Rev
10:453476, 1990
Fischer KW, Ayoub C: Affective splitting and dissociation
in normal and maltreated children: developmental
pathways for self in relationships, in Disorders and
Dysfunctions of the Self: Rochester Symposium on Developmental Psychopathology, Vol 5. Edited by Cicchetti D, Toth SL. New York, Plenum Press, 1994, pp
149222

184

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

Fonagy P, Target M, Gergely G: Attachment and borderline


personality disorder: a theory and some evidence. Psychiatr Clin North Am 23:103122, 2000
Fonagy P, Target M, Gergely G, et al: The developmental
roots of borderline personality disorder in early attachment relationships: a theory and some evidence. Psychoanalytic Inquiry 23:412459, 2003
Geiger TC, Crick NR: A developmental psychopathology
perspective on vulnerability to personality disorders, in
Vulnerability to Psychopathology. Edited by Ingram
RE, Price JM. New York, Guilford, 2001, pp 57102
Golombek H, Marton P, Stein B, et al: Personality functioning status during early and middle adolescence. Adolesc Psychiatry 14:365377, 1987
Grilo CM, Becker DF, Fehon DC, et al: Gender differences in
personality disorders in psychiatrically hospitalized adolescents. Am J Psychiatry 153:10891091, 1996
Harter S: The development of self-representations, in Handbook of Child Psychology, Vol 3: Social, Emotional, and
Personality Development, 5th Edition. Edited by Damon
W, Eisenberg N. New York, Wiley, 1998, pp 553617
Johnson JG, Cohen P, Skodol AE, et al: Personality disorders
in adolescence and risk of major mental disorders and
suicidality during adulthood. Arch Gen Psychiatry 56:
805811, 1999
Johnson JG, Cohen P, Kasen S, et al: Age-related change in
personality disorder trait levels between early adolescence and adulthood: a community-based longitudinal
investigation. Acta Psychiatr Scand 102:265275, 2000a
Johnson JG, Cohen P, Smailes E, et al: Adolescent personality
disorders associated with violence and criminal behavior during adolescence and early adulthood. Am J Psychiatry 157:14061412, 2000b
Kasen S, Cohen P, Skodol AE, et al: Influence of child and adolescent psychiatric disorders on young adult personality disorder. Am J Psychiatry 156:15291535, 1999
Kasen S, Cohen P, Skodol AE, et al: Childhood depression
and adult personality disorder: alternative pathways of
continuity. Arch Gen Psychiatry 58:231236, 2001
Keating D: Adolescent thinking, in At the Threshold: The Developing Adolescent. Edited by Feldman SS, Elliott G.
Cambridge, MA, Harvard University Press, 1990, pp 5490
Kernberg OF: Borderline Conditions and Pathological Narcissism. Northvale, NJ, Jason Aronson, 1975
Kernberg OF, Weiner AS, Bardenstein KK: Personality Disorders in Children and Adolescents. New York, Basic
Books, 2000
Klonsky ED, Oltmanns TF, Turkheirmer E: Informant reports of personality disorders: relation to self-reports
and future research directions. Clinical Psychology: Science and Practice 9:300311, 2002
Kobak R, Cole C: Attachment and metamonitoring: implications for adolescent autonomy and psychopathology, in
Rochester Symposium on Development and Psychopathology, Vol 5: Disorders of the Self. Edited by Cicchetti
D. Rochester, NY, University of Rochester Press, 1994,
pp 267297

Lachar D: Personality Inventory for Children-2 (PIC-2) Manual. Los Angeles, CA, Western Psychological Services,
1999a
Lachar D: Personality Inventory for Children, 2nd Edition
(PIC-2), Personality Inventory for Youth (PIY), and Student Behavior Survey (SBS), in The Use of Psychological
Testing for Treatment Planning and Outcomes Assessment, 2nd Edition. Edited by Maruish ME. Mahwah,
NJ, Lawrence Erlbaum Associates, 1999b, pp 399427
Lachar D, Gruber CP: Personality Inventory for Youth (PIY)
Manual. Los Angeles, CA, Western Psychological Services, 1995
Lewinsohn PM, Rohde P, Seeley JR, et al: Axis II psychopathology as a function of Axis I disorders in childhood
and adolescence. J Am Acad Child Adolesc Psychiatry
36:17521759, 1997
Linehan MM: Cognitive Behavioral Treatment of Borderline
Personality Disorder. New York, Guilford, 1993
Livesley WJ, Jang KL, Vernon PA: The phenotypic and genetic architecture of traits delineating personality disorder. Arch Gen Psychiatry 55:941948, 1998
Lofgren DP, Bemporad J, King J, et al: A follow-up study of socalled borderline children. Am J Psychiatry 148:1541
1547, 1991
Loranger AW: Personality Disorder Examination (personality disorder E) Manual. Yonkers, NY, DV Communications, 1988
MacCallum RC, Zhang S, Preacher KJ, et al: On the practice
of dichotomization of quantitative variables. Psychol
Methods 7:4163, 2002
Marton P, Golombek H, Stein B, et al: Behavior disturbance
and changes in personality dysfunction from early to
middle adolescence. Adolesc Psychiatry 14:394406,
1987
Mattanah JJF, Becker DF, Levy KN, et al: Diagnostic stability
in adolescents followed 2 years after hospitalization.
Am J Psychiatry 152:889894, 1995
McCrae RR, Costa PT, de Lima MP, et al: Age differences in
personality across the adult life span: parallels in five
cultures. Dev Psychol 35:466477, 1999
McCrae RR, Costa PT, Terracciano A, et al: Personality trait
development from age 12 to age 18: longitudinal, crosssectional, and cross-cultural analyses. J Pers Soc Psychol
83:14561468, 2002
McDermott PA: A nationwide study of developmental and
gender prevalence for psychopathology in childhood
and adolescence. J Abnorm Child Psychol 24:5366,
1996
Meeus W, Iedema J, Helsen M, et al: Patterns of adolescent
identity development: review of literature and longitudinal analysis. Dev Rev 19:419461, 1999
Millon T: Toward a New Personology: An Evolutionary
Model. New York, Wiley, 1990
Moffitt T: Adolescence-limited and life-course-persistent antisocial behavior: a developmental taxonomy. Psychol
Rev 100:674701, 1993

Developmental Issues

Pine F: On the development of the borderline-child-to-be.


Am J Orthopsychiatry 56:450457, 1986
Pinto A, Grapentine WL, Francis G, et al: Borderline personality disorder in adolescents: affective and cognitive
features. J Am Acad Child Adolesc Psychiatry 35:1338
1343, 1996
Rey JM, Singh M, Morris-Yates A, et al: Referred adolescents
as young adults: the relationship between psychosocial
functioning and personality disorder. Aust N Z J Psychiatry 31:219226, 1997
Reynolds WM: Adolescent Psychopathology Scale. Odessa,
FL, Psychological Assessment Resources, 1998
Roberts BW, DelVecchio WF: The rank-order consistency of
personality traits from childhood to old age: a quantitative review of longitudinal studies. Psychol Bull 126:3
25, 2000
Rothbard JC, Shaver PR: Continuity of attachment across the
life span, in Attachment in Adults: Clinical and Developmental Perspectives. Edited by Sperling MB, Berman
WH. New York, Guilford, 1994, pp 3171
Rothbart MK, Bates JE: Temperament, in Handbook of Child
Psychology, 5th Edition, Vol 3: Social, Emotional, and
Personality Development. Edited by Damon M, Eisenberg N. New York, Wiley, 1998, pp 105176
Rothbart MK, Ahadi SA, Hershey KL, et al: Investigations of
temperament at three to seven years: the Childrens Behavior Questionnaire. Child Dev 72:13941408, 2001
Shaffer D, Fisher P, Lucas CP, et al: NIMH Diagnostic Interview Schedule for Children Version IV (NIMH DISCIV): description, differences from previous versions,
and reliability of some common diagnoses. J Am Acad
Child Adolesc Psychiatry 39:2838, 2000
Shea MT, Stout R, Gunderson J, et al: Short-term diagnostic
stability of schizotypal, borderline, avoidant, and obsessive-compulsive personality disorders. Am J Psychiatry
159:20362041, 2002

185

Shedler J, Westen D: Refining the measurement of Axis II: a


Q-sort procedure for assessing personality pathology.
Assessment 5:335355, 1998
Shiner R, Caspi A: Personality differences in childhood and
adolescence: measurement, development, and consequences. J Child Psychol Psychiatry 44:232, 2003
Siegler RS: Childrens Thinking, 2nd Edition. Englewood
Cliffs, NJ, Prentice Hall, 1991
Solomon J, George C: Attachment Disorganization. New
York, Guilford, 1999
Sroufe LA: Emotional Development: The Organization of
Emotional Life in the Early Years. New York, Cambridge University Press, 1996
Stein B, Golombek H, Marton P, et al: Personality functioning and change in clinical presentation from early to
middle adolescence. Adolesc Psychiatry 14:378393,
1987
Taylor S, Goritsas E: Dimensions of identity diffusion. J Personal Disord 8:229239, 1994
Tellegen A: Structure of mood and personality and their relevance to assessing anxiety, with an emphasis on selfreport, in Anxiety and the Anxiety Disorders. Edited by
Tuma AH, Maser JD. Hillsdale, NJ, Erlbaum, 1985, pp
681706
Waters E, Merrick S, Treboux D, et al: Attachment security in
infancy and early adulthood: a longitudinal study.
Child Dev 71:684689, 2000
Westen D, Shedler J, Glass S, et al: Personality diagnoses in
adolescence: DSM-IV Axis II diagnoses and an empirically derived alternative. Am J Psychiatry 160:952966,
2003
Widiger TA, Sankis L: Adult psychopathology: issues and
controversies. Ann Rev Psychol 51:377404, 2000
Zanarini MC, Frankenburg FR, Hennen J, et al: The longitudinal course of borderline psychopathology: 6-year prospective follow-up of the phenomenology of borderline
personality disorder. Am J Psychiatry 160:274283, 2003

This page intentionally left blank

12
Attachment Theory and
Mentalization-Oriented
Model of Borderline
Personality Disorder
Peter Fonagy, Ph.D., F.B.A.
Anthony W. Bateman, M.A., F.R.C.Psych.

Borderline personality disorder (BPD) is a dysfunction of self-regulation particularly in the context of social relationships. Both the regulation of emotion and
the catastrophic reaction to the loss of intensely emotionally invested social ties place BPD in the domain of
attachment. A number of theorists have drawn on
Bowlbys ideas in explanation of borderline pathology. Most specifically, Gunderson (1996) suggested
that intolerance of aloneness was at the core of borderline pathology, and the inability of those with BPD to
invoke a soothing introject was a consequence of
early attachment failures. He carefully described typical patterns of borderline dysfunction in terms of exaggerated reactions of the insecurely attached infant;
for example, clinging, fearfulness about dependency
needs, terror of abandonment, and constant monitor-

ing of the proximity of the caregiver. Lyons-Ruth and


Jacobovitz (1999) focused on the disorganization of
the attachment system in infancy as predisposing to
later borderline pathology. Notably, they identified an
insecure, as opposed to a secure, disorganized pattern
as predisposing to conduct problems. Crittenden
(1997) was particularly concerned with incorporating
borderline individuals deep ambivalence and fear of
close relationships in her representation of adult attachment disorganization. Fonagy and colleagues
(Fonagy 2000; Fonagy et al. 2000) also used the framework of attachment theory but emphasized the role of
attachment in the development of symbolic function
and the way in which insecure disorganized attachment may generate vulnerability in the face of further
turmoil and challenges. All of these, and other, theo187

188

T e x tb o o k o f P e r s o n a l i ty D i s o r d e r s

retical approaches predict that representations of attachment will be seriously insecure and arguably disorganized in patients with BPD.
In this chapter we briefly outline the theory of attachment and some empirical work linking BPD with
dysfunctions of the attachment system. We consider
BPD from an attachment theory perspective, introducing some modifications of classical attachment theory
that have helped us understand the disordered attachment of individuals with BPD. In particular, we link
the development of the capacity of mentalization (the
ability to represent the behavior of self and others in
terms of underlying mental states) with the quality of
attachment relationshipsand link the failure of mentalization with symptoms of BPD. Finally, the treatment implications of our attachment theorybased
model of BPD are discussed.

BRIEF OUTLINE OF ATTACHMENT THEORY


Bowlbys attachment theory has a biological focus
(Bowlby 1969). Attachment readily reduces to a molecular level of infant behaviors, such as smiling and
vocalizing, that alert the caregiver to the childs interest in socializing and bring the caregiver close to the
child. Smiling and vocalizing are attachment behaviors, as is crying, which is experienced by most caregivers as aversive, and they engage the caregiver in
caretaking behaviors. Bowlby emphasized the survival value of attachment in enhancing safety through
proximity to the caregiver in addition to feeding,
learning about the environment, and social interaction, as well as protection from predators. Bowlby
(1969) considered the latter to be the biological function of attachment behavior. Attachment behaviors
were seen as part of a behavioral system (a term
Bowlby borrowed from ethology).
In the second volume of his Attachment and Loss
trilogy, Bowlby established the set goal of the attachment system as maintaining the caregivers accessibility and responsiveness, which he covered with a single
term: availability (Bowlby 1973). Availability means confident expectationgained from tolerably accurately (p. 202) represented experience over a significant time periodthat the attachment figure will be
available. The attachment behavioral system thus
came to be underpinned by a set of cognitive mechanisms, discussed by Bowlby as representational models or by Craik (1943) as internal working models
(Bretherton and Munholland 1999; Crittenden 1994;
Main 1991; Sroufe 1996). Four representational sys-

tems are implied by the internal working models:


1) expectations of interactive attributes of early caregivers created in the first year of life and subsequently
elaborated; 2) event representations by which general
and specific memories of attachment-related experiences are encoded and retrieved; 3) autobiographical
memories by which specific events are conceptually
connected because of their relation to a continuing personal narrative and developing self-understanding;
and 4) understanding of the psychological characteristics of other people and differentiating them from the
characteristics of the self. It is in this last layer of the internal working models that we consider the dysfunctions of individuals with BPD to be most profound.
The second great pioneer of attachment theory,
Mary Ainsworth (1969, 1985; Ainsworth et al. 1978),
developed the well-known laboratory-based procedure of the Strange Situation for observing infants internal working models in action. When infants are
briefly separated from their caregivers in an unfamiliar
situation, they show one of four patterns of behavior.
Infants who display secure attachment explore readily
in the presence of the caregiver, are anxious in the presence of the stranger and avoid her, are distressed by
the caregivers brief absence, rapidly seek contact with
the caregiver afterward, and are reassured by this contact and return to their exploration. Some infants, designated as anxious/avoidant, appear to be made less
anxious by separation, may not seek contact with the
caregiver following separation, and may not prefer her
over the stranger. Anxious/resistant infants show limited exploration and play, tend to be highly distressed
by separation from the caregiver, and have great difficulty in settling afterward, showing struggling, stiffness, continued crying, or fuss in a passive way. The
caregivers presence or attempts at comforting fail to
reassure, and the infants anxiety and anger appear to
prevent him from deriving comfort from proximity.
A fourth group of infants who show seemingly undirected behavior are referred to as disorganized/disoriented (Main and Solomon 1990). They show freezing,
hand clapping, head banging, and a wish to escape the
situation even in the presence of the caregiver (LyonsRuth and Jacobovitz 1999; Van IJzendoorn et al. 1999).
It is generally held that for such infants the caregiver
has served as a source of both fear and reassurance,
and thus arousal of the attachment behavioral system
produces strong conflicting motivations.
Prospective longitudinal research has demonstrated that children with a history of secure attachment are independently rated as more resilient, selfreliant, socially oriented (Sroufe 1983; Waters et al.

Attachment Theory and Mentalization-Oriented Model of Borderline Personality Disorder

1979), and empathic to distress (Kestenbaum et al.


1989), with deeper relationships and higher selfesteem (Sroufe 1983; Sroufe et al. 1990). Bowlby proposed that internal working models of the self and
others provide prototypes for all later relationships.
Such models are relatively stable across the lifespan
(Collins and Read 1994).
Because internal working models function outside
of awareness, they are change resistant (Crittenden
1990). The stability of attachment is demonstrated by
longitudinal studies of infants assessed with the
strange situation and followed up in adolescence or
young adulthood with the Adult Attachment Interview
(AAI; George C, Kaplan N, Main M: The Adult Attachment Interview. Unpublished manuscript, Department of Psychology, University of California at Berkeley, 1985). This structured clinical instrument elicits
narrative histories of childhood attachment relationshipsthe characteristics of early relationships, experiences of separation, illness, punishment, loss, maltreatment, or abuse. The AAI scoring system (Main M,
Goldwyn R: Adult Attachment Rating and Classification System, Manual in Draft, Version 6.0. Unpublished manuscript, University of California at Berkeley,
1994) classifies individuals into secure/autonomous,
insecure/dismissing, insecure/preoccupied, or unresolved with respect to loss or trauma, which are categories based on the structural qualities of narratives of
early experiences. Whereas autonomous individuals
value attachment relationships, coherently integrate
memories into a meaningful narrative, and regard these
as formative, insecure individuals are poor at integrating memories of experience with the meaning of that
experience. Those individuals who are dismissing of attachment show avoidance by denying memories and
by idealizing or devaluing (or both idealizing and devaluing) early relationships. Preoccupied individuals
tend to be confused, angry, or passive in relation to attachment figures, often still complaining of childhood
slights, echoing the protests of the resistant infant. Unresolved individuals give indications of significant disorganization in their attachment relationship representation; this disorganization manifests in semantic or
syntactic confusions in their narratives concerning
childhood trauma or a recent loss.
Many studies have demonstrated that the AAI, administered to the mother or father, will predict not
only the childs security of attachment to that parent
but even more remarkably the precise attachment category that the child manifests in the strange situation
(Van IJzendoorn 1995). Thus, a dismissing AAI interview predicts avoidant strange-situation behavior,

189

whereas a preoccupied interview predicts anxious/resistant infant attachment. Lack of resolution of mourning (unresolved interviews) predicts disorganization
in infant attachment (discussed later). Temperament
(child-to-parent effects) seems an inadequate account
of the phenomena, because t